Author: "Guez, Arthur" - Searchworks@Jio Institute Digital Library Search Results

Your search keyword '"Guez, Arthur"' showing total 39 results

Start Over Author "Guez, Arthur"

39 results on '"Guez, Arthur"'

1. A Unifying Framework for Action-Conditional Self-Predictive Reinforcement Learning

Author: Khetarpal, Khimya, Guo, Zhaohan Daniel, Pires, Bernardo Avila, Tang, Yunhao, Lyle, Clare, Rowland, Mark, Heess, Nicolas, Borsa, Diana, Guez, Arthur, and Dabney, Will
Subjects: Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract: Learning a good representation is a crucial challenge for Reinforcement Learning (RL) agents. Self-predictive learning provides means to jointly learn a latent representation and dynamics model by bootstrapping from future latent representations (BYOL). Recent work has developed theoretical insights into these algorithms by studying a continuous-time ODE model for self-predictive representation learning under the simplifying assumption that the algorithm depends on a fixed policy (BYOL-$\Pi$); this assumption is at odds with practical instantiations of such algorithms, which explicitly condition their predictions on future actions. In this work, we take a step towards bridging the gap between theory and practice by analyzing an action-conditional self-predictive objective (BYOL-AC) using the ODE framework, characterizing its convergence properties and highlighting important distinctions between the limiting solutions of the BYOL-$\Pi$ and BYOL-AC dynamics. We show how the two representations are related by a variance equation. This connection leads to a novel variance-like action-conditional objective (BYOL-VAR) and its corresponding ODE. We unify the study of all three objectives through two complementary lenses; a model-based perspective, where each objective is shown to be equivalent to a low-rank approximation of certain dynamics, and a model-free perspective, which establishes relationships between the objectives and their respective value, Q-value, and advantage function. Our empirical investigations, encompassing both linear function approximation and Deep RL environments, demonstrates that BYOL-AC is better overall in a variety of different settings.
Published: 2024

2. Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context

Author: Gemini Team, Georgiev, Petko, Lei, Ving Ian, Burnell, Ryan, Bai, Libin, Gulati, Anmol, Tanzer, Garrett, Vincent, Damien, Pan, Zhufeng, Wang, Shibo, Mariooryad, Soroosh, Ding, Yifan, Geng, Xinyang, Alcober, Fred, Frostig, Roy, Omernick, Mark, Walker, Lexi, Paduraru, Cosmin, Sorokin, Christina, Tacchetti, Andrea, Gaffney, Colin, Daruki, Samira, Sercinoglu, Olcan, Gleicher, Zach, Love, Juliette, Voigtlaender, Paul, Jain, Rohan, Surita, Gabriela, Mohamed, Kareem, Blevins, Rory, Ahn, Junwhan, Zhu, Tao, Kawintiranon, Kornraphop, Firat, Orhan, Gu, Yiming, Zhang, Yujing, Rahtz, Matthew, Faruqui, Manaal, Clay, Natalie, Gilmer, Justin, Co-Reyes, JD, Penchev, Ivo, Zhu, Rui, Morioka, Nobuyuki, Hui, Kevin, Haridasan, Krishna, Campos, Victor, Mahdieh, Mahdis, Guo, Mandy, Hassan, Samer, Kilgour, Kevin, Vezer, Arpi, Cheng, Heng-Tze, de Liedekerke, Raoul, Goyal, Siddharth, Barham, Paul, Strouse, DJ, Noury, Seb, Adler, Jonas, Sundararajan, Mukund, Vikram, Sharad, Lepikhin, Dmitry, Paganini, Michela, Garcia, Xavier, Yang, Fan, Valter, Dasha, Trebacz, Maja, Vodrahalli, Kiran, Asawaroengchai, Chulayuth, Ring, Roman, Kalb, Norbert, Soares, Livio Baldini, Brahma, Siddhartha, Steiner, David, Yu, Tianhe, Mentzer, Fabian, He, Antoine, Gonzalez, Lucas, Xu, Bibo, Kaufman, Raphael Lopez, Shafey, Laurent El, Oh, Junhyuk, Hennigan, Tom, Driessche, George van den, Odoom, Seth, Lucic, Mario, Roelofs, Becca, Lall, Sid, Marathe, Amit, Chan, Betty, Ontanon, Santiago, He, Luheng, Teplyashin, Denis, Lai, Jonathan, Crone, Phil, Damoc, Bogdan, Ho, Lewis, Riedel, Sebastian, Lenc, Karel, Yeh, Chih-Kuan, Chowdhery, Aakanksha, Xu, Yang, Kazemi, Mehran, Amid, Ehsan, Petrushkina, Anastasia, Swersky, Kevin, Khodaei, Ali, Chen, Gowoon, Larkin, Chris, Pinto, Mario, Yan, Geng, Badia, Adria Puigdomenech, Patil, Piyush, Hansen, Steven, Orr, Dave, Arnold, Sebastien M. R., Grimstad, Jordan, Dai, Andrew, Douglas, Sholto, Sinha, Rishika, Yadav, Vikas, Chen, Xi, Gribovskaya, Elena, Austin, Jacob, Zhao, Jeffrey, Patel, Kaushal, Komarek, Paul, Austin, Sophia, Borgeaud, Sebastian, Friso, Linda, Goyal, Abhimanyu, Caine, Ben, Cao, Kris, Chung, Da-Woon, Lamm, Matthew, Barth-Maron, Gabe, Kagohara, Thais, Olszewska, Kate, Chen, Mia, Shivakumar, Kaushik, Agarwal, Rishabh, Godhia, Harshal, Rajwar, Ravi, Snaider, Javier, Dotiwalla, Xerxes, Liu, Yuan, Barua, Aditya, Ungureanu, Victor, Zhang, Yuan, Batsaikhan, Bat-Orgil, Wirth, Mateo, Qin, James, Danihelka, Ivo, Doshi, Tulsee, Chadwick, Martin, Chen, Jilin, Jain, Sanil, Le, Quoc, Kar, Arjun, Gurumurthy, Madhu, Li, Cheng, Sang, Ruoxin, Liu, Fangyu, Lamprou, Lampros, Munoz, Rich, Lintz, Nathan, Mehta, Harsh, Howard, Heidi, Reynolds, Malcolm, Aroyo, Lora, Wang, Quan, Blanco, Lorenzo, Cassirer, Albin, Griffith, Jordan, Das, Dipanjan, Lee, Stephan, Sygnowski, Jakub, Fisher, Zach, Besley, James, Powell, Richard, Ahmed, Zafarali, Paulus, Dominik, Reitter, David, Borsos, Zalan, Joshi, Rishabh, Pope, Aedan, Hand, Steven, Selo, Vittorio, Jain, Vihan, Sethi, Nikhil, Goel, Megha, Makino, Takaki, May, Rhys, Yang, Zhen, Schalkwyk, Johan, Butterfield, Christina, Hauth, Anja, Goldin, Alex, Hawkins, Will, Senter, Evan, Brin, Sergey, Woodman, Oliver, Ritter, Marvin, Noland, Eric, Giang, Minh, Bolina, Vijay, Lee, Lisa, Blyth, Tim, Mackinnon, Ian, Reid, Machel, Sarvana, Obaid, Silver, David, Chen, Alexander, Wang, Lily, Maggiore, Loren, Chang, Oscar, Attaluri, Nithya, Thornton, Gregory, Chiu, Chung-Cheng, Bunyan, Oskar, Levine, Nir, Chung, Timothy, Eltyshev, Evgenii, Si, Xiance, Lillicrap, Timothy, Brady, Demetra, Aggarwal, Vaibhav, Wu, Boxi, Xu, Yuanzhong, McIlroy, Ross, Badola, Kartikeya, Sandhu, Paramjit, Moreira, Erica, Stokowiec, Wojciech, Hemsley, Ross, Li, Dong, Tudor, Alex, Shyam, Pranav, Rahimtoroghi, Elahe, Haykal, Salem, Sprechmann, Pablo, Zhou, Xiang, Mincu, Diana, Li, Yujia, Addanki, Ravi, Krishna, Kalpesh, Wu, Xiao, Frechette, Alexandre, Eyal, Matan, Dafoe, Allan, Lacey, Dave, Whang, Jay, Avrahami, Thi, Zhang, Ye, Taropa, Emanuel, Lin, Hanzhao, Toyama, Daniel, Rutherford, Eliza, Sano, Motoki, Choe, HyunJeong, Tomala, Alex, Safranek-Shrader, Chalence, Kassner, Nora, Pajarskas, Mantas, Harvey, Matt, Sechrist, Sean, Fortunato, Meire, Lyu, Christina, Elsayed, Gamaleldin, Kuang, Chenkai, Lottes, James, Chu, Eric, Jia, Chao, Chen, Chih-Wei, Humphreys, Peter, Baumli, Kate, Tao, Connie, Samuel, Rajkumar, Santos, Cicero Nogueira dos, Andreassen, Anders, Rakićević, Nemanja, Grewe, Dominik, Kumar, Aviral, Winkler, Stephanie, Caton, Jonathan, Brock, Andrew, Dalmia, Sid, Sheahan, Hannah, Barr, Iain, Miao, Yingjie, Natsev, Paul, Devlin, Jacob, Behbahani, Feryal, Prost, Flavien, Sun, Yanhua, Myaskovsky, Artiom, Pillai, Thanumalayan Sankaranarayana, Hurt, Dan, Lazaridou, Angeliki, Xiong, Xi, Zheng, Ce, Pardo, Fabio, Li, Xiaowei, Horgan, Dan, Stanton, Joe, Ambar, Moran, Xia, Fei, Lince, Alejandro, Wang, Mingqiu, Mustafa, Basil, Webson, Albert, Lee, Hyo, Anil, Rohan, Wicke, Martin, Dozat, Timothy, Sinha, Abhishek, Piqueras, Enrique, Dabir, Elahe, Upadhyay, Shyam, Boral, Anudhyan, Hendricks, Lisa Anne, Fry, Corey, Djolonga, Josip, Su, Yi, Walker, Jake, Labanowski, Jane, Huang, Ronny, Misra, Vedant, Chen, Jeremy, Skerry-Ryan, RJ, Singh, Avi, Rijhwani, Shruti, Yu, Dian, Castro-Ros, Alex, Changpinyo, Beer, Datta, Romina, Bagri, Sumit, Hrafnkelsson, Arnar Mar, Maggioni, Marcello, Zheng, Daniel, Sulsky, Yury, Hou, Shaobo, Paine, Tom Le, Yang, Antoine, Riesa, Jason, Rogozinska, Dominika, Marcus, Dror, Badawy, Dalia El, Zhang, Qiao, Wang, Luyu, Miller, Helen, Greer, Jeremy, Sjos, Lars Lowe, Nova, Azade, Zen, Heiga, Chaabouni, Rahma, Rosca, Mihaela, Jiang, Jiepu, Chen, Charlie, Liu, Ruibo, Sainath, Tara, Krikun, Maxim, Polozov, Alex, Lespiau, Jean-Baptiste, Newlan, Josh, Cankara, Zeyncep, Kwak, Soo, Xu, Yunhan, Chen, Phil, Coenen, Andy, Meyer, Clemens, Tsihlas, Katerina, Ma, Ada, Gottweis, Juraj, Xing, Jinwei, Gu, Chenjie, Miao, Jin, Frank, Christian, Cankara, Zeynep, Ganapathy, Sanjay, Dasgupta, Ishita, Hughes-Fitt, Steph, Chen, Heng, Reid, David, Rong, Keran, Fan, Hongmin, van Amersfoort, Joost, Zhuang, Vincent, Cohen, Aaron, Gu, Shixiang Shane, Mohananey, Anhad, Ilic, Anastasija, Tobin, Taylor, Wieting, John, Bortsova, Anna, Thacker, Phoebe, Wang, Emma, Caveness, Emily, Chiu, Justin, Sezener, Eren, Kaskasoli, Alex, Baker, Steven, Millican, Katie, Elhawaty, Mohamed, Aisopos, Kostas, Lebsack, Carl, Byrd, Nathan, Dai, Hanjun, Jia, Wenhao, Wiethoff, Matthew, Davoodi, Elnaz, Weston, Albert, Yagati, Lakshman, Ahuja, Arun, Gao, Isabel, Pundak, Golan, Zhang, Susan, Azzam, Michael, Sim, Khe Chai, Caelles, Sergi, Keeling, James, Sharma, Abhanshu, Swing, Andy, Li, YaGuang, Liu, Chenxi, Bostock, Carrie Grimes, Bansal, Yamini, Nado, Zachary, Anand, Ankesh, Lipschultz, Josh, Karmarkar, Abhijit, Proleev, Lev, Ittycheriah, Abe, Yeganeh, Soheil Hassas, Polovets, George, Faust, Aleksandra, Sun, Jiao, Rrustemi, Alban, Li, Pen, Shivanna, Rakesh, Liu, Jeremiah, Welty, Chris, Lebron, Federico, Baddepudi, Anirudh, Krause, Sebastian, Parisotto, Emilio, Soricut, Radu, Xu, Zheng, Bloxwich, Dawn, Johnson, Melvin, Neyshabur, Behnam, Mao-Jones, Justin, Wang, Renshen, Ramasesh, Vinay, Abbas, Zaheer, Guez, Arthur, Segal, Constant, Nguyen, Duc Dung, Svensson, James, Hou, Le, York, Sarah, Milan, Kieran, Bridgers, Sophie, Gworek, Wiktor, Tagliasacchi, Marco, Lee-Thorp, James, Chang, Michael, Guseynov, Alexey, Hartman, Ale Jakse, Kwong, Michael, Zhao, Ruizhe, Kashem, Sheleem, Cole, Elizabeth, Miech, Antoine, Tanburn, Richard, Phuong, Mary, Pavetic, Filip, Cevey, Sebastien, Comanescu, Ramona, Ives, Richard, Yang, Sherry, Du, Cosmo, Li, Bo, Zhang, Zizhao, Iinuma, Mariko, Hu, Clara Huiyi, Roy, Aurko, Bijwadia, Shaan, Zhu, Zhenkai, Martins, Danilo, Saputro, Rachel, Gergely, Anita, Zheng, Steven, Jia, Dawei, Antonoglou, Ioannis, Sadovsky, Adam, Gu, Shane, Bi, Yingying, Andreev, Alek, Samangooei, Sina, Khan, Mina, Kocisky, Tomas, Filos, Angelos, Kumar, Chintu, Bishop, Colton, Yu, Adams, Hodkinson, Sarah, Mittal, Sid, Shah, Premal, Moufarek, Alexandre, Cheng, Yong, Bloniarz, Adam, Lee, Jaehoon, Pejman, Pedram, Michel, Paul, Spencer, Stephen, Feinberg, Vladimir, Xiong, Xuehan, Savinov, Nikolay, Smith, Charlotte, Shakeri, Siamak, Tran, Dustin, Chesus, Mary, Bohnet, Bernd, Tucker, George, von Glehn, Tamara, Muir, Carrie, Mao, Yiran, Kazawa, Hideto, Slone, Ambrose, Soparkar, Kedar, Shrivastava, Disha, Cobon-Kerr, James, Sharman, Michael, Pavagadhi, Jay, Araya, Carlos, Misiunas, Karolis, Ghelani, Nimesh, Laskin, Michael, Barker, David, Li, Qiujia, Briukhov, Anton, Houlsby, Neil, Glaese, Mia, Lakshminarayanan, Balaji, Schucher, Nathan, Tang, Yunhao, Collins, Eli, Lim, Hyeontaek, Feng, Fangxiaoyu, Recasens, Adria, Lai, Guangda, Magni, Alberto, De Cao, Nicola, Siddhant, Aditya, Ashwood, Zoe, Orbay, Jordi, Dehghani, Mostafa, Brennan, Jenny, He, Yifan, Xu, Kelvin, Gao, Yang, Saroufim, Carl, Molloy, James, Wu, Xinyi, Arnold, Seb, Chang, Solomon, Schrittwieser, Julian, Buchatskaya, Elena, Radpour, Soroush, Polacek, Martin, Giordano, Skye, Bapna, Ankur, Tokumine, Simon, Hellendoorn, Vincent, Sottiaux, Thibault, Cogan, Sarah, Severyn, Aliaksei, Saleh, Mohammad, Thakoor, Shantanu, Shefey, Laurent, Qiao, Siyuan, Gaba, Meenu, Chang, Shuo-yiin, Swanson, Craig, Zhang, Biao, Lee, Benjamin, Rubenstein, Paul Kishan, Song, Gan, Kwiatkowski, Tom, Koop, Anna, Kannan, Ajay, Kao, David, Schuh, Parker, Stjerngren, Axel, Ghiasi, Golnaz, Gibson, Gena, Vilnis, Luke, Yuan, Ye, Ferreira, Felipe Tiengo, Kamath, Aishwarya, Klimenko, Ted, Franko, Ken, Xiao, Kefan, Bhattacharya, Indro, Patel, Miteyan, Wang, Rui, Morris, Alex, Strudel, Robin, Sharma, Vivek, Choy, Peter, Hashemi, Sayed Hadi, Landon, Jessica, Finkelstein, Mara, Jhakra, Priya, Frye, Justin, Barnes, Megan, Mauger, Matthew, Daun, Dennis, Baatarsukh, Khuslen, Tung, Matthew, Farhan, Wael, Michalewski, Henryk, Viola, Fabio, Quitry, Felix de Chaumont, Lan, Charline Le, Hudson, Tom, Wang, Qingze, Fischer, Felix, Zheng, Ivy, White, Elspeth, Dragan, Anca, Alayrac, Jean-baptiste, Ni, Eric, Pritzel, Alexander, Iwanicki, Adam, Isard, Michael, Bulanova, Anna, Zilka, Lukas, Dyer, Ethan, Sachan, Devendra, Srinivasan, Srivatsan, Muckenhirn, Hannah, Cai, Honglong, Mandhane, Amol, Tariq, Mukarram, Rae, Jack W., Wang, Gary, Ayoub, Kareem, FitzGerald, Nicholas, Zhao, Yao, Han, Woohyun, Alberti, Chris, Garrette, Dan, Krishnakumar, Kashyap, Gimenez, Mai, Levskaya, Anselm, Sohn, Daniel, Matak, Josip, Iturrate, Inaki, Chang, Michael B., Xiang, Jackie, Cao, Yuan, Ranka, Nishant, Brown, Geoff, Hutter, Adrian, Mirrokni, Vahab, Chen, Nanxin, Yao, Kaisheng, Egyed, Zoltan, Galilee, Francois, Liechty, Tyler, Kallakuri, Praveen, Palmer, Evan, Ghemawat, Sanjay, Liu, Jasmine, Tao, David, Thornton, Chloe, Green, Tim, Jasarevic, Mimi, Lin, Sharon, Cotruta, Victor, Tan, Yi-Xuan, Fiedel, Noah, Yu, Hongkun, Chi, Ed, Neitz, Alexander, Heitkaemper, Jens, Sinha, Anu, Zhou, Denny, Sun, Yi, Kaed, Charbel, Hulse, Brice, Mishra, Swaroop, Georgaki, Maria, Kudugunta, Sneha, Farabet, Clement, Shafran, Izhak, Vlasic, Daniel, Tsitsulin, Anton, Ananthanarayanan, Rajagopal, Carin, Alen, Su, Guolong, Sun, Pei, V, Shashank, Carvajal, Gabriel, Broder, Josef, Comsa, Iulia, Repina, Alena, Wong, William, Chen, Warren Weilun, Hawkins, Peter, Filonov, Egor, Loher, Lucia, Hirnschall, Christoph, Wang, Weiyi, Ye, Jingchen, Burns, Andrea, Cate, Hardie, Wright, Diana Gage, Piccinini, Federico, Zhang, Lei, Lin, Chu-Cheng, Gog, Ionel, Kulizhskaya, Yana, Sreevatsa, Ashwin, Song, Shuang, Cobo, Luis C., Iyer, Anand, Tekur, Chetan, Garrido, Guillermo, Xiao, Zhuyun, Kemp, Rupert, Zheng, Huaixiu Steven, Li, Hui, Agarwal, Ananth, Ngani, Christel, Goshvadi, Kati, Santamaria-Fernandez, Rebeca, Fica, Wojciech, Chen, Xinyun, Gorgolewski, Chris, Sun, Sean, Garg, Roopal, Ye, Xinyu, Eslami, S. M. Ali, Hua, Nan, Simon, Jon, Joshi, Pratik, Kim, Yelin, Tenney, Ian, Potluri, Sahitya, Thiet, Lam Nguyen, Yuan, Quan, Luisier, Florian, Chronopoulou, Alexandra, Scellato, Salvatore, Srinivasan, Praveen, Chen, Minmin, Koverkathu, Vinod, Dalibard, Valentin, Xu, Yaming, Saeta, Brennan, Anderson, Keith, Sellam, Thibault, Fernando, Nick, Huot, Fantine, Jung, Junehyuk, Varadarajan, Mani, Quinn, Michael, Raul, Amit, Le, Maigo, Habalov, Ruslan, Clark, Jon, Jalan, Komal, Bullard, Kalesha, Singhal, Achintya, Luong, Thang, Wang, Boyu, Rajayogam, Sujeevan, Eisenschlos, Julian, Jia, Johnson, Finchelstein, Daniel, Yakubovich, Alex, Balle, Daniel, Fink, Michael, Agarwal, Sameer, Li, Jing, Dvijotham, Dj, Pal, Shalini, Kang, Kai, Konzelmann, Jaclyn, Beattie, Jennifer, Dousse, Olivier, Wu, Diane, Crocker, Remi, Elkind, Chen, Jonnalagadda, Siddhartha Reddy, Lee, Jong, Holtmann-Rice, Dan, Kallarackal, Krystal, Liu, Rosanne, Vnukov, Denis, Vats, Neera, Invernizzi, Luca, Jafari, Mohsen, Zhou, Huanjie, Taylor, Lilly, Prendki, Jennifer, Wu, Marcus, Eccles, Tom, Liu, Tianqi, Kopparapu, Kavya, Beaufays, Francoise, Angermueller, Christof, Marzoca, Andreea, Sarcar, Shourya, Dib, Hilal, Stanway, Jeff, Perbet, Frank, Trdin, Nejc, Sterneck, Rachel, Khorlin, Andrey, Li, Dinghua, Wu, Xihui, Goenka, Sonam, Madras, David, Goldshtein, Sasha, Gierke, Willi, Zhou, Tong, Liu, Yaxin, Liang, Yannie, White, Anais, Li, Yunjie, Singh, Shreya, Bahargam, Sanaz, Epstein, Mark, Basu, Sujoy, Lao, Li, Ozturel, Adnan, Crous, Carl, Zhai, Alex, Lu, Han, Tung, Zora, Gaur, Neeraj, Walton, Alanna, Dixon, Lucas, Zhang, Ming, Globerson, Amir, Uy, Grant, Bolt, Andrew, Wiles, Olivia, Nasr, Milad, Shumailov, Ilia, Selvi, Marco, Piccinno, Francesco, Aguilar, Ricardo, McCarthy, Sara, Khalman, Misha, Shukla, Mrinal, Galic, Vlado, Carpenter, John, Villela, Kevin, Zhang, Haibin, Richardson, Harry, Martens, James, Bosnjak, Matko, Belle, Shreyas Rammohan, Seibert, Jeff, Alnahlawi, Mahmoud, McWilliams, Brian, Singh, Sankalp, Louis, Annie, Ding, Wen, Popovici, Dan, Simicich, Lenin, Knight, Laura, Mehta, Pulkit, Gupta, Nishesh, Shi, Chongyang, Fatehi, Saaber, Mitrovic, Jovana, Grills, Alex, Pagadora, Joseph, Munkhdalai, Tsendsuren, Petrova, Dessie, Eisenbud, Danielle, Zhang, Zhishuai, Yates, Damion, Mittal, Bhavishya, Tripuraneni, Nilesh, Assael, Yannis, Brovelli, Thomas, Jain, Prateek, Velimirovic, Mihajlo, Akbulut, Canfer, Mu, Jiaqi, Macherey, Wolfgang, Kumar, Ravin, Xu, Jun, Qureshi, Haroon, Comanici, Gheorghe, Wiesner, Jeremy, Gong, Zhitao, Ruddock, Anton, Bauer, Matthias, Felt, Nick, GP, Anirudh, Arnab, Anurag, Zelle, Dustin, Rothfuss, Jonas, Rosgen, Bill, Shenoy, Ashish, Seybold, Bryan, Li, Xinjian, Mudigonda, Jayaram, Erdogan, Goker, Xia, Jiawei, Simsa, Jiri, Michi, Andrea, Yao, Yi, Yew, Christopher, Kan, Steven, Caswell, Isaac, Radebaugh, Carey, Elisseeff, Andre, Valenzuela, Pedro, McKinney, Kay, Paterson, Kim, Cui, Albert, Latorre-Chimoto, Eri, Kim, Solomon, Zeng, William, Durden, Ken, Ponnapalli, Priya, Sosea, Tiberiu, Choquette-Choo, Christopher A., Manyika, James, Robenek, Brona, Vashisht, Harsha, Pereira, Sebastien, Lam, Hoi, Velic, Marko, Owusu-Afriyie, Denese, Lee, Katherine, Bolukbasi, Tolga, Parrish, Alicia, Lu, Shawn, Park, Jane, Venkatraman, Balaji, Talbert, Alice, Rosique, Lambert, Cheng, Yuchung, Sozanschi, Andrei, Paszke, Adam, Kumar, Praveen, Austin, Jessica, Li, Lu, Salama, Khalid, Perz, Bartek, Kim, Wooyeol, Dukkipati, Nandita, Baryshnikov, Anthony, Kaplanis, Christos, Sheng, XiangHai, Chervonyi, Yuri, Unlu, Caglar, Casas, Diego de Las, Askham, Harry, Tunyasuvunakool, Kathryn, Gimeno, Felix, Poder, Siim, Kwak, Chester, Miecnikowski, Matt, Dimitriev, Alek, Parisi, Aaron, Liu, Dangyi, Tsai, Tomy, Shevlane, Toby, Kouridi, Christina, Garmon, Drew, Goedeckemeyer, Adrian, Brown, Adam R., Vijayakumar, Anitha, Elqursh, Ali, Jazayeri, Sadegh, Huang, Jin, Carthy, Sara Mc, Hoover, Jay, Kim, Lucy, Kumar, Sandeep, Chen, Wei, Biles, Courtney, Bingham, Garrett, Rosen, Evan, Wang, Lisa, Tan, Qijun, Engel, David, Pongetti, Francesco, de Cesare, Dario, Hwang, Dongseong, Yu, Lily, Pullman, Jennifer, Narayanan, Srini, Levin, Kyle, Gopal, Siddharth, Li, Megan, Aharoni, Asaf, Trinh, Trieu, Lo, Jessica, Casagrande, Norman, Vij, Roopali, Matthey, Loic, Ramadhana, Bramandia, Matthews, Austin, Carey, CJ, Johnson, Matthew, Goranova, Kremena, Shah, Rohin, Ashraf, Shereen, Dasgupta, Kingshuk, Larsen, Rasmus, Wang, Yicheng, Vuyyuru, Manish Reddy, Jiang, Chong, Ijazi, Joana, Osawa, Kazuki, Smith, Celine, Boppana, Ramya Sree, Bilal, Taylan, Koizumi, Yuma, Xu, Ying, Altun, Yasemin, Shabat, Nir, Bariach, Ben, Korchemniy, Alex, Choo, Kiam, Ronneberger, Olaf, Iwuanyanwu, Chimezie, Zhao, Shubin, Soergel, David, Hsieh, Cho-Jui, Cai, Irene, Iqbal, Shariq, Sundermeyer, Martin, Chen, Zhe, Bursztein, Elie, Malaviya, Chaitanya, Biadsy, Fadi, Shroff, Prakash, Dhillon, Inderjit, Latkar, Tejasi, Dyer, Chris, Forbes, Hannah, Nicosia, Massimo, Nikolaev, Vitaly, Greene, Somer, Georgiev, Marin, Wang, Pidong, Martin, Nina, Sedghi, Hanie, Zhang, John, Banzal, Praseem, Fritz, Doug, Rao, Vikram, Wang, Xuezhi, Zhang, Jiageng, Patraucean, Viorica, Du, Dayou, Mordatch, Igor, Jurin, Ivan, Liu, Lewis, Dubey, Ayush, Mohan, Abhi, Nowakowski, Janek, Ion, Vlad-Doru, Wei, Nan, Tojo, Reiko, Raad, Maria Abi, Hudson, Drew A., Keshava, Vaishakh, Agrawal, Shubham, Ramirez, Kevin, Wu, Zhichun, Nguyen, Hoang, Liu, Ji, Sewak, Madhavi, Petrini, Bryce, Choi, DongHyun, Philips, Ivan, Wang, Ziyue, Bica, Ioana, Garg, Ankush, Wilkiewicz, Jarek, Agrawal, Priyanka, Guo, Danhao, Xue, Emily, Shaik, Naseer, Leach, Andrew, Khan, Sadh MNM, Wiesinger, Julia, Jerome, Sammy, Chakladar, Abhishek, Wang, Alek Wenjiao, Ornduff, Tina, Abu, Folake, Ghaffarkhah, Alireza, Wainwright, Marcus, Cortes, Mario, Liu, Frederick, Maynez, Joshua, Terzis, Andreas, Samangouei, Pouya, Mansour, Riham, Kępa, Tomasz, Aubet, François-Xavier, Algymr, Anton, Banica, Dan, Weisz, Agoston, Orban, Andras, Senges, Alexandre, Andrejczuk, Ewa, Geller, Mark, Santo, Niccolo Dal, Anklin, Valentin, Merey, Majd Al, Baeuml, Martin, Strohman, Trevor, Bai, Junwen, Petrov, Slav, Wu, Yonghui, Hassabis, Demis, Kavukcuoglu, Koray, Dean, Jeff, and Vinyals, Oriol
Subjects: Computer Science - Computation and Language, Computer Science - Artificial Intelligence
Abstract: In this report, we introduce the Gemini 1.5 family of models, representing the next generation of highly compute-efficient multimodal models capable of recalling and reasoning over fine-grained information from millions of tokens of context, including multiple long documents and hours of video and audio. The family includes two new models: (1) an updated Gemini 1.5 Pro, which exceeds the February version on the great majority of capabilities and benchmarks; (2) Gemini 1.5 Flash, a more lightweight variant designed for efficiency with minimal regression in quality. Gemini 1.5 models achieve near-perfect recall on long-context retrieval tasks across modalities, improve the state-of-the-art in long-document QA, long-video QA and long-context ASR, and match or surpass Gemini 1.0 Ultra's state-of-the-art performance across a broad set of benchmarks. Studying the limits of Gemini 1.5's long-context ability, we find continued improvement in next-token prediction and near-perfect retrieval (>99%) up to at least 10M tokens, a generational leap over existing models such as Claude 3.0 (200k) and GPT-4 Turbo (128k). Finally, we highlight real-world use cases, such as Gemini 1.5 collaborating with professionals on completing their tasks achieving 26 to 75% time savings across 10 different job categories, as well as surprising new capabilities of large language models at the frontier; when given a grammar manual for Kalamang, a language with fewer than 200 speakers worldwide, the model learns to translate English to Kalamang at a similar level to a person who learned from the same content.
Published: 2024

3. Gemini: A Family of Highly Capable Multimodal Models

Author: Gemini Team, Anil, Rohan, Borgeaud, Sebastian, Alayrac, Jean-Baptiste, Yu, Jiahui, Soricut, Radu, Schalkwyk, Johan, Dai, Andrew M., Hauth, Anja, Millican, Katie, Silver, David, Johnson, Melvin, Antonoglou, Ioannis, Schrittwieser, Julian, Glaese, Amelia, Chen, Jilin, Pitler, Emily, Lillicrap, Timothy, Lazaridou, Angeliki, Firat, Orhan, Molloy, James, Isard, Michael, Barham, Paul R., Hennigan, Tom, Lee, Benjamin, Viola, Fabio, Reynolds, Malcolm, Xu, Yuanzhong, Doherty, Ryan, Collins, Eli, Meyer, Clemens, Rutherford, Eliza, Moreira, Erica, Ayoub, Kareem, Goel, Megha, Krawczyk, Jack, Du, Cosmo, Chi, Ed, Cheng, Heng-Tze, Ni, Eric, Shah, Purvi, Kane, Patrick, Chan, Betty, Faruqui, Manaal, Severyn, Aliaksei, Lin, Hanzhao, Li, YaGuang, Cheng, Yong, Ittycheriah, Abe, Mahdieh, Mahdis, Chen, Mia, Sun, Pei, Tran, Dustin, Bagri, Sumit, Lakshminarayanan, Balaji, Liu, Jeremiah, Orban, Andras, Güra, Fabian, Zhou, Hao, Song, Xinying, Boffy, Aurelien, Ganapathy, Harish, Zheng, Steven, Choe, HyunJeong, Weisz, Ágoston, Zhu, Tao, Lu, Yifeng, Gopal, Siddharth, Kahn, Jarrod, Kula, Maciej, Pitman, Jeff, Shah, Rushin, Taropa, Emanuel, Merey, Majd Al, Baeuml, Martin, Chen, Zhifeng, Shafey, Laurent El, Zhang, Yujing, Sercinoglu, Olcan, Tucker, George, Piqueras, Enrique, Krikun, Maxim, Barr, Iain, Savinov, Nikolay, Danihelka, Ivo, Roelofs, Becca, White, Anaïs, Andreassen, Anders, von Glehn, Tamara, Yagati, Lakshman, Kazemi, Mehran, Gonzalez, Lucas, Khalman, Misha, Sygnowski, Jakub, Frechette, Alexandre, Smith, Charlotte, Culp, Laura, Proleev, Lev, Luan, Yi, Chen, Xi, Lottes, James, Schucher, Nathan, Lebron, Federico, Rrustemi, Alban, Clay, Natalie, Crone, Phil, Kocisky, Tomas, Zhao, Jeffrey, Perz, Bartek, Yu, Dian, Howard, Heidi, Bloniarz, Adam, Rae, Jack W., Lu, Han, Sifre, Laurent, Maggioni, Marcello, Alcober, Fred, Garrette, Dan, Barnes, Megan, Thakoor, Shantanu, Austin, Jacob, Barth-Maron, Gabriel, Wong, William, Joshi, Rishabh, Chaabouni, Rahma, Fatiha, Deeni, Ahuja, Arun, Tomar, Gaurav Singh, Senter, Evan, Chadwick, Martin, Kornakov, Ilya, Attaluri, Nithya, Iturrate, Iñaki, Liu, Ruibo, Li, Yunxuan, Cogan, Sarah, Chen, Jeremy, Jia, Chao, Gu, Chenjie, Zhang, Qiao, Grimstad, Jordan, Hartman, Ale Jakse, Garcia, Xavier, Pillai, Thanumalayan Sankaranarayana, Devlin, Jacob, Laskin, Michael, Casas, Diego de Las, Valter, Dasha, Tao, Connie, Blanco, Lorenzo, Badia, Adrià Puigdomènech, Reitter, David, Chen, Mianna, Brennan, Jenny, Rivera, Clara, Brin, Sergey, Iqbal, Shariq, Surita, Gabriela, Labanowski, Jane, Rao, Abhi, Winkler, Stephanie, Parisotto, Emilio, Gu, Yiming, Olszewska, Kate, Addanki, Ravi, Miech, Antoine, Louis, Annie, Teplyashin, Denis, Brown, Geoff, Catt, Elliot, Balaguer, Jan, Xiang, Jackie, Wang, Pidong, Ashwood, Zoe, Briukhov, Anton, Webson, Albert, Ganapathy, Sanjay, Sanghavi, Smit, Kannan, Ajay, Chang, Ming-Wei, Stjerngren, Axel, Djolonga, Josip, Sun, Yuting, Bapna, Ankur, Aitchison, Matthew, Pejman, Pedram, Michalewski, Henryk, Yu, Tianhe, Wang, Cindy, Love, Juliette, Ahn, Junwhan, Bloxwich, Dawn, Han, Kehang, Humphreys, Peter, Sellam, Thibault, Bradbury, James, Godbole, Varun, Samangooei, Sina, Damoc, Bogdan, Kaskasoli, Alex, Arnold, Sébastien M. R., Vasudevan, Vijay, Agrawal, Shubham, Riesa, Jason, Lepikhin, Dmitry, Tanburn, Richard, Srinivasan, Srivatsan, Lim, Hyeontaek, Hodkinson, Sarah, Shyam, Pranav, Ferret, Johan, Hand, Steven, Garg, Ankush, Paine, Tom Le, Li, Jian, Li, Yujia, Giang, Minh, Neitz, Alexander, Abbas, Zaheer, York, Sarah, Reid, Machel, Cole, Elizabeth, Chowdhery, Aakanksha, Das, Dipanjan, Rogozińska, Dominika, Nikolaev, Vitaliy, Sprechmann, Pablo, Nado, Zachary, Zilka, Lukas, Prost, Flavien, He, Luheng, Monteiro, Marianne, Mishra, Gaurav, Welty, Chris, Newlan, Josh, Jia, Dawei, Allamanis, Miltiadis, Hu, Clara Huiyi, de Liedekerke, Raoul, Gilmer, Justin, Saroufim, Carl, Rijhwani, Shruti, Hou, Shaobo, Shrivastava, Disha, Baddepudi, Anirudh, Goldin, Alex, Ozturel, Adnan, Cassirer, Albin, Xu, Yunhan, Sohn, Daniel, Sachan, Devendra, Amplayo, Reinald Kim, Swanson, Craig, Petrova, Dessie, Narayan, Shashi, Guez, Arthur, Brahma, Siddhartha, Landon, Jessica, Patel, Miteyan, Zhao, Ruizhe, Villela, Kevin, Wang, Luyu, Jia, Wenhao, Rahtz, Matthew, Giménez, Mai, Yeung, Legg, Keeling, James, Georgiev, Petko, Mincu, Diana, Wu, Boxi, Haykal, Salem, Saputro, Rachel, Vodrahalli, Kiran, Qin, James, Cankara, Zeynep, Sharma, Abhanshu, Fernando, Nick, Hawkins, Will, Neyshabur, Behnam, Kim, Solomon, Hutter, Adrian, Agrawal, Priyanka, Castro-Ros, Alex, Driessche, George van den, Wang, Tao, Yang, Fan, Chang, Shuo-yiin, Komarek, Paul, McIlroy, Ross, Lučić, Mario, Zhang, Guodong, Farhan, Wael, Sharman, Michael, Natsev, Paul, Michel, Paul, Bansal, Yamini, Qiao, Siyuan, Cao, Kris, Shakeri, Siamak, Butterfield, Christina, Chung, Justin, Rubenstein, Paul Kishan, Agrawal, Shivani, Mensch, Arthur, Soparkar, Kedar, Lenc, Karel, Chung, Timothy, Pope, Aedan, Maggiore, Loren, Kay, Jackie, Jhakra, Priya, Wang, Shibo, Maynez, Joshua, Phuong, Mary, Tobin, Taylor, Tacchetti, Andrea, Trebacz, Maja, Robinson, Kevin, Katariya, Yash, Riedel, Sebastian, Bailey, Paige, Xiao, Kefan, Ghelani, Nimesh, Aroyo, Lora, Slone, Ambrose, Houlsby, Neil, Xiong, Xuehan, Yang, Zhen, Gribovskaya, Elena, Adler, Jonas, Wirth, Mateo, Lee, Lisa, Li, Music, Kagohara, Thais, Pavagadhi, Jay, Bridgers, Sophie, Bortsova, Anna, Ghemawat, Sanjay, Ahmed, Zafarali, Liu, Tianqi, Powell, Richard, Bolina, Vijay, Iinuma, Mariko, Zablotskaia, Polina, Besley, James, Chung, Da-Woon, Dozat, Timothy, Comanescu, Ramona, Si, Xiance, Greer, Jeremy, Su, Guolong, Polacek, Martin, Kaufman, Raphaël Lopez, Tokumine, Simon, Hu, Hexiang, Buchatskaya, Elena, Miao, Yingjie, Elhawaty, Mohamed, Siddhant, Aditya, Tomasev, Nenad, Xing, Jinwei, Greer, Christina, Miller, Helen, Ashraf, Shereen, Roy, Aurko, Zhang, Zizhao, Ma, Ada, Filos, Angelos, Besta, Milos, Blevins, Rory, Klimenko, Ted, Yeh, Chih-Kuan, Changpinyo, Soravit, Mu, Jiaqi, Chang, Oscar, Pajarskas, Mantas, Muir, Carrie, Cohen, Vered, Lan, Charline Le, Haridasan, Krishna, Marathe, Amit, Hansen, Steven, Douglas, Sholto, Samuel, Rajkumar, Wang, Mingqiu, Austin, Sophia, Lan, Chang, Jiang, Jiepu, Chiu, Justin, Lorenzo, Jaime Alonso, Sjösund, Lars Lowe, Cevey, Sébastien, Gleicher, Zach, Avrahami, Thi, Boral, Anudhyan, Srinivasan, Hansa, Selo, Vittorio, May, Rhys, Aisopos, Konstantinos, Hussenot, Léonard, Soares, Livio Baldini, Baumli, Kate, Chang, Michael B., Recasens, Adrià, Caine, Ben, Pritzel, Alexander, Pavetic, Filip, Pardo, Fabio, Gergely, Anita, Frye, Justin, Ramasesh, Vinay, Horgan, Dan, Badola, Kartikeya, Kassner, Nora, Roy, Subhrajit, Dyer, Ethan, Campos, Víctor Campos, Tomala, Alex, Tang, Yunhao, Badawy, Dalia El, White, Elspeth, Mustafa, Basil, Lang, Oran, Jindal, Abhishek, Vikram, Sharad, Gong, Zhitao, Caelles, Sergi, Hemsley, Ross, Thornton, Gregory, Feng, Fangxiaoyu, Stokowiec, Wojciech, Zheng, Ce, Thacker, Phoebe, Ünlü, Çağlar, Zhang, Zhishuai, Saleh, Mohammad, Svensson, James, Bileschi, Max, Patil, Piyush, Anand, Ankesh, Ring, Roman, Tsihlas, Katerina, Vezer, Arpi, Selvi, Marco, Shevlane, Toby, Rodriguez, Mikel, Kwiatkowski, Tom, Daruki, Samira, Rong, Keran, Dafoe, Allan, FitzGerald, Nicholas, Gu-Lemberg, Keren, Khan, Mina, Hendricks, Lisa Anne, Pellat, Marie, Feinberg, Vladimir, Cobon-Kerr, James, Sainath, Tara, Rauh, Maribeth, Hashemi, Sayed Hadi, Ives, Richard, Hasson, Yana, Noland, Eric, Cao, Yuan, Byrd, Nathan, Hou, Le, Wang, Qingze, Sottiaux, Thibault, Paganini, Michela, Lespiau, Jean-Baptiste, Moufarek, Alexandre, Hassan, Samer, Shivakumar, Kaushik, van Amersfoort, Joost, Mandhane, Amol, Joshi, Pratik, Goyal, Anirudh, Tung, Matthew, Brock, Andrew, Sheahan, Hannah, Misra, Vedant, Li, Cheng, Rakićević, Nemanja, Dehghani, Mostafa, Liu, Fangyu, Mittal, Sid, Oh, Junhyuk, Noury, Seb, Sezener, Eren, Huot, Fantine, Lamm, Matthew, De Cao, Nicola, Chen, Charlie, Mudgal, Sidharth, Stella, Romina, Brooks, Kevin, Vasudevan, Gautam, Liu, Chenxi, Chain, Mainak, Melinkeri, Nivedita, Cohen, Aaron, Wang, Venus, Seymore, Kristie, Zubkov, Sergey, Goel, Rahul, Yue, Summer, Krishnakumaran, Sai, Albert, Brian, Hurley, Nate, Sano, Motoki, Mohananey, Anhad, Joughin, Jonah, Filonov, Egor, Kępa, Tomasz, Eldawy, Yomna, Lim, Jiawern, Rishi, Rahul, Badiezadegan, Shirin, Bos, Taylor, Chang, Jerry, Jain, Sanil, Padmanabhan, Sri Gayatri Sundara, Puttagunta, Subha, Krishna, Kalpesh, Baker, Leslie, Kalb, Norbert, Bedapudi, Vamsi, Kurzrok, Adam, Lei, Shuntong, Yu, Anthony, Litvin, Oren, Zhou, Xiang, Wu, Zhichun, Sobell, Sam, Siciliano, Andrea, Papir, Alan, Neale, Robby, Bragagnolo, Jonas, Toor, Tej, Chen, Tina, Anklin, Valentin, Wang, Feiran, Feng, Richie, Gholami, Milad, Ling, Kevin, Liu, Lijuan, Walter, Jules, Moghaddam, Hamid, Kishore, Arun, Adamek, Jakub, Mercado, Tyler, Mallinson, Jonathan, Wandekar, Siddhinita, Cagle, Stephen, Ofek, Eran, Garrido, Guillermo, Lombriser, Clemens, Mukha, Maksim, Sun, Botu, Mohammad, Hafeezul Rahman, Matak, Josip, Qian, Yadi, Peswani, Vikas, Janus, Pawel, Yuan, Quan, Schelin, Leif, David, Oana, Garg, Ankur, He, Yifan, Duzhyi, Oleksii, Älgmyr, Anton, Lottaz, Timothée, Li, Qi, Yadav, Vikas, Xu, Luyao, Chinien, Alex, Shivanna, Rakesh, Chuklin, Aleksandr, Li, Josie, Spadine, Carrie, Wolfe, Travis, Mohamed, Kareem, Das, Subhabrata, Dai, Zihang, He, Kyle, von Dincklage, Daniel, Upadhyay, Shyam, Maurya, Akanksha, Chi, Luyan, Krause, Sebastian, Salama, Khalid, Rabinovitch, Pam G, M, Pavan Kumar Reddy, Selvan, Aarush, Dektiarev, Mikhail, Ghiasi, Golnaz, Guven, Erdem, Gupta, Himanshu, Liu, Boyi, Sharma, Deepak, Shtacher, Idan Heimlich, Paul, Shachi, Akerlund, Oscar, Aubet, François-Xavier, Huang, Terry, Zhu, Chen, Zhu, Eric, Teixeira, Elico, Fritze, Matthew, Bertolini, Francesco, Marinescu, Liana-Eleonora, Bölle, Martin, Paulus, Dominik, Gupta, Khyatti, Latkar, Tejasi, Chang, Max, Sanders, Jason, Wilson, Roopa, Wu, Xuewei, Tan, Yi-Xuan, Thiet, Lam Nguyen, Doshi, Tulsee, Lall, Sid, Mishra, Swaroop, Chen, Wanming, Luong, Thang, Benjamin, Seth, Lee, Jasmine, Andrejczuk, Ewa, Rabiej, Dominik, Ranjan, Vipul, Styrc, Krzysztof, Yin, Pengcheng, Simon, Jon, Harriott, Malcolm Rose, Bansal, Mudit, Robsky, Alexei, Bacon, Geoff, Greene, David, Mirylenka, Daniil, Zhou, Chen, Sarvana, Obaid, Goyal, Abhimanyu, Andermatt, Samuel, Siegler, Patrick, Horn, Ben, Israel, Assaf, Pongetti, Francesco, Chen, Chih-Wei "Louis", Selvatici, Marco, Silva, Pedro, Wang, Kathie, Tolins, Jackson, Guu, Kelvin, Yogev, Roey, Cai, Xiaochen, Agostini, Alessandro, Shah, Maulik, Nguyen, Hung, Donnaile, Noah Ó, Pereira, Sébastien, Friso, Linda, Stambler, Adam, Kuang, Chenkai, Romanikhin, Yan, Geller, Mark, Yan, ZJ, Jang, Kane, Lee, Cheng-Chun, Fica, Wojciech, Malmi, Eric, Tan, Qijun, Banica, Dan, Balle, Daniel, Pham, Ryan, Huang, Yanping, Avram, Diana, Shi, Hongzhi, Singh, Jasjot, Hidey, Chris, Ahuja, Niharika, Saxena, Pranab, Dooley, Dan, Potharaju, Srividya Pranavi, O'Neill, Eileen, Gokulchandran, Anand, Foley, Ryan, Zhao, Kai, Dusenberry, Mike, Liu, Yuan, Mehta, Pulkit, Kotikalapudi, Ragha, Safranek-Shrader, Chalence, Goodman, Andrew, Kessinger, Joshua, Globen, Eran, Kolhar, Prateek, Gorgolewski, Chris, Ibrahim, Ali, Song, Yang, Eichenbaum, Ali, Brovelli, Thomas, Potluri, Sahitya, Lahoti, Preethi, Baetu, Cip, Ghorbani, Ali, Chen, Charles, Crawford, Andy, Pal, Shalini, Sridhar, Mukund, Gurita, Petru, Mujika, Asier, Petrovski, Igor, Cedoz, Pierre-Louis, Li, Chenmei, Chen, Shiyuan, Santo, Niccolò Dal, Goyal, Siddharth, Punjabi, Jitesh, Kappaganthu, Karthik, Kwak, Chester, LV, Pallavi, Velury, Sarmishta, Choudhury, Himadri, Hall, Jamie, Shah, Premal, Figueira, Ricardo, Thomas, Matt, Lu, Minjie, Zhou, Ting, Kumar, Chintu, Jurdi, Thomas, Chikkerur, Sharat, Ma, Yenai, Yu, Adams, Kwak, Soo, Ähdel, Victor, Rajayogam, Sujeevan, Choma, Travis, Liu, Fei, Barua, Aditya, Ji, Colin, Park, Ji Ho, Hellendoorn, Vincent, Bailey, Alex, Bilal, Taylan, Zhou, Huanjie, Khatir, Mehrdad, Sutton, Charles, Rzadkowski, Wojciech, Macintosh, Fiona, Shagin, Konstantin, Medina, Paul, Liang, Chen, Zhou, Jinjing, Shah, Pararth, Bi, Yingying, Dankovics, Attila, Banga, Shipra, Lehmann, Sabine, Bredesen, Marissa, Lin, Zifan, Hoffmann, John Eric, Lai, Jonathan, Chung, Raynald, Yang, Kai, Balani, Nihal, Bražinskas, Arthur, Sozanschi, Andrei, Hayes, Matthew, Alcalde, Héctor Fernández, Makarov, Peter, Chen, Will, Stella, Antonio, Snijders, Liselotte, Mandl, Michael, Kärrman, Ante, Nowak, Paweł, Wu, Xinyi, Dyck, Alex, Vaidyanathan, Krishnan, R, Raghavender, Mallet, Jessica, Rudominer, Mitch, Johnston, Eric, Mittal, Sushil, Udathu, Akhil, Christensen, Janara, Verma, Vishal, Irving, Zach, Santucci, Andreas, Elsayed, Gamaleldin, Davoodi, Elnaz, Georgiev, Marin, Tenney, Ian, Hua, Nan, Cideron, Geoffrey, Leurent, Edouard, Alnahlawi, Mahmoud, Georgescu, Ionut, Wei, Nan, Zheng, Ivy, Scandinaro, Dylan, Jiang, Heinrich, Snoek, Jasper, Sundararajan, Mukund, Wang, Xuezhi, Ontiveros, Zack, Karo, Itay, Cole, Jeremy, Rajashekhar, Vinu, Tumeh, Lara, Ben-David, Eyal, Jain, Rishub, Uesato, Jonathan, Datta, Romina, Bunyan, Oskar, Wu, Shimu, Zhang, John, Stanczyk, Piotr, Zhang, Ye, Steiner, David, Naskar, Subhajit, Azzam, Michael, Johnson, Matthew, Paszke, Adam, Chiu, Chung-Cheng, Elias, Jaume Sanchez, Mohiuddin, Afroz, Muhammad, Faizan, Miao, Jin, Lee, Andrew, Vieillard, Nino, Park, Jane, Zhang, Jiageng, Stanway, Jeff, Garmon, Drew, Karmarkar, Abhijit, Dong, Zhe, Lee, Jong, Kumar, Aviral, Zhou, Luowei, Evens, Jonathan, Isaac, William, Irving, Geoffrey, Loper, Edward, Fink, Michael, Arkatkar, Isha, Chen, Nanxin, Shafran, Izhak, Petrychenko, Ivan, Chen, Zhe, Jia, Johnson, Levskaya, Anselm, Zhu, Zhenkai, Grabowski, Peter, Mao, Yu, Magni, Alberto, Yao, Kaisheng, Snaider, Javier, Casagrande, Norman, Palmer, Evan, Suganthan, Paul, Castaño, Alfonso, Giannoumis, Irene, Kim, Wooyeol, Rybiński, Mikołaj, Sreevatsa, Ashwin, Prendki, Jennifer, Soergel, David, Goedeckemeyer, Adrian, Gierke, Willi, Jafari, Mohsen, Gaba, Meenu, Wiesner, Jeremy, Wright, Diana Gage, Wei, Yawen, Vashisht, Harsha, Kulizhskaya, Yana, Hoover, Jay, Le, Maigo, Li, Lu, Iwuanyanwu, Chimezie, Liu, Lu, Ramirez, Kevin, Khorlin, Andrey, Cui, Albert, LIN, Tian, Wu, Marcus, Aguilar, Ricardo, Pallo, Keith, Chakladar, Abhishek, Perng, Ginger, Abellan, Elena Allica, Zhang, Mingyang, Dasgupta, Ishita, Kushman, Nate, Penchev, Ivo, Repina, Alena, Wu, Xihui, van der Weide, Tom, Ponnapalli, Priya, Kaplan, Caroline, Simsa, Jiri, Li, Shuangfeng, Dousse, Olivier, Piper, Jeff, Ie, Nathan, Pasumarthi, Rama, Lintz, Nathan, Vijayakumar, Anitha, Andor, Daniel, Valenzuela, Pedro, Lui, Minnie, Paduraru, Cosmin, Peng, Daiyi, Lee, Katherine, Zhang, Shuyuan, Greene, Somer, Nguyen, Duc Dung, Kurylowicz, Paula, Hardin, Cassidy, Dixon, Lucas, Janzer, Lili, Choo, Kiam, Feng, Ziqiang, Zhang, Biao, Singhal, Achintya, Du, Dayou, McKinnon, Dan, Antropova, Natasha, Bolukbasi, Tolga, Keller, Orgad, Reid, David, Finchelstein, Daniel, Raad, Maria Abi, Crocker, Remi, Hawkins, Peter, Dadashi, Robert, Gaffney, Colin, Franko, Ken, Bulanova, Anna, Leblond, Rémi, Chung, Shirley, Askham, Harry, Cobo, Luis C., Xu, Kelvin, Fischer, Felix, Xu, Jun, Sorokin, Christina, Alberti, Chris, Lin, Chu-Cheng, Evans, Colin, Dimitriev, Alek, Forbes, Hannah, Banarse, Dylan, Tung, Zora, Omernick, Mark, Bishop, Colton, Sterneck, Rachel, Jain, Rohan, Xia, Jiawei, Amid, Ehsan, Piccinno, Francesco, Wang, Xingyu, Banzal, Praseem, Mankowitz, Daniel J., Polozov, Alex, Krakovna, Victoria, Brown, Sasha, Bateni, MohammadHossein, Duan, Dennis, Firoiu, Vlad, Thotakuri, Meghana, Natan, Tom, Geist, Matthieu, Girgin, Ser tan, Li, Hui, Ye, Jiayu, Roval, Ofir, Tojo, Reiko, Kwong, Michael, Lee-Thorp, James, Yew, Christopher, Sinopalnikov, Danila, Ramos, Sabela, Mellor, John, Sharma, Abhishek, Wu, Kathy, Miller, David, Sonnerat, Nicolas, Vnukov, Denis, Greig, Rory, Beattie, Jennifer, Caveness, Emily, Bai, Libin, Eisenschlos, Julian, Korchemniy, Alex, Tsai, Tomy, Jasarevic, Mimi, Kong, Weize, Dao, Phuong, Zheng, Zeyu, Liu, Frederick, Zhu, Rui, Teh, Tian Huey, Sanmiya, Jason, Gladchenko, Evgeny, Trdin, Nejc, Toyama, Daniel, Rosen, Evan, Tavakkol, Sasan, Xue, Linting, Elkind, Chen, Woodman, Oliver, Carpenter, John, Papamakarios, George, Kemp, Rupert, Kafle, Sushant, Grunina, Tanya, Sinha, Rishika, Talbert, Alice, Wu, Diane, Owusu-Afriyie, Denese, Thornton, Chloe, Pont-Tuset, Jordi, Narayana, Pradyumna, Li, Jing, Fatehi, Saaber, Wieting, John, Ajmeri, Omar, Uria, Benigno, Ko, Yeongil, Knight, Laura, Héliou, Amélie, Niu, Ning, Gu, Shane, Pang, Chenxi, Li, Yeqing, Levine, Nir, Stolovich, Ariel, Santamaria-Fernandez, Rebeca, Goenka, Sonam, Yustalim, Wenny, Strudel, Robin, Elqursh, Ali, Deck, Charlie, Lee, Hyo, Li, Zonglin, Levin, Kyle, Hoffmann, Raphael, Holtmann-Rice, Dan, Bachem, Olivier, Arora, Sho, Koh, Christy, Yeganeh, Soheil Hassas, Põder, Siim, Tariq, Mukarram, Sun, Yanhua, Ionita, Lucian, Seyedhosseini, Mojtaba, Tafti, Pouya, Liu, Zhiyu, Gulati, Anmol, Liu, Jasmine, Ye, Xinyu, Chrzaszcz, Bart, Wang, Lily, Sethi, Nikhil, Li, Tianrun, Brown, Ben, Singh, Shreya, Fan, Wei, Parisi, Aaron, Stanton, Joe, Koverkathu, Vinod, Choquette-Choo, Christopher A., Li, Yunjie, Lu, TJ, Shroff, Prakash, Varadarajan, Mani, Bahargam, Sanaz, Willoughby, Rob, Gaddy, David, Desjardins, Guillaume, Cornero, Marco, Robenek, Brona, Mittal, Bhavishya, Albrecht, Ben, Shenoy, Ashish, Moiseev, Fedor, Jacobsson, Henrik, Ghaffarkhah, Alireza, Rivière, Morgane, Walton, Alanna, Crepy, Clément, Parrish, Alicia, Zhou, Zongwei, Farabet, Clement, Radebaugh, Carey, Srinivasan, Praveen, van der Salm, Claudia, Fidjeland, Andreas, Scellato, Salvatore, Latorre-Chimoto, Eri, Klimczak-Plucińska, Hanna, Bridson, David, de Cesare, Dario, Hudson, Tom, Mendolicchio, Piermaria, Walker, Lexi, Morris, Alex, Mauger, Matthew, Guseynov, Alexey, Reid, Alison, Odoom, Seth, Loher, Lucia, Cotruta, Victor, Yenugula, Madhavi, Grewe, Dominik, Petrushkina, Anastasia, Duerig, Tom, Sanchez, Antonio, Yadlowsky, Steve, Shen, Amy, Globerson, Amir, Webb, Lynette, Dua, Sahil, Li, Dong, Bhupatiraju, Surya, Hurt, Dan, Qureshi, Haroon, Agarwal, Ananth, Shani, Tomer, Eyal, Matan, Khare, Anuj, Belle, Shreyas Rammohan, Wang, Lei, Tekur, Chetan, Kale, Mihir Sanjay, Wei, Jinliang, Sang, Ruoxin, Saeta, Brennan, Liechty, Tyler, Sun, Yi, Zhao, Yao, Lee, Stephan, Nayak, Pandu, Fritz, Doug, Vuyyuru, Manish Reddy, Aslanides, John, Vyas, Nidhi, Wicke, Martin, Ma, Xiao, Eltyshev, Evgenii, Martin, Nina, Cate, Hardie, Manyika, James, Amiri, Keyvan, Kim, Yelin, Xiong, Xi, Kang, Kai, Luisier, Florian, Tripuraneni, Nilesh, Madras, David, Guo, Mandy, Waters, Austin, Wang, Oliver, Ainslie, Joshua, Baldridge, Jason, Zhang, Han, Pruthi, Garima, Bauer, Jakob, Yang, Feng, Mansour, Riham, Gelman, Jason, Xu, Yang, Polovets, George, Liu, Ji, Cai, Honglong, Chen, Warren, Sheng, XiangHai, Xue, Emily, Ozair, Sherjil, Angermueller, Christof, Li, Xiaowei, Sinha, Anoop, Wang, Weiren, Wiesinger, Julia, Koukoumidis, Emmanouil, Tian, Yuan, Iyer, Anand, Gurumurthy, Madhu, Goldenson, Mark, Shah, Parashar, Blake, MK, Yu, Hongkun, Urbanowicz, Anthony, Palomaki, Jennimaria, Fernando, Chrisantha, Durden, Ken, Mehta, Harsh, Momchev, Nikola, Rahimtoroghi, Elahe, Georgaki, Maria, Raul, Amit, Ruder, Sebastian, Redshaw, Morgan, Lee, Jinhyuk, Zhou, Denny, Jalan, Komal, Li, Dinghua, Hechtman, Blake, Schuh, Parker, Nasr, Milad, Milan, Kieran, Mikulik, Vladimir, Franco, Juliana, Green, Tim, Nguyen, Nam, Kelley, Joe, Mahendru, Aroma, Hu, Andrea, Howland, Joshua, Vargas, Ben, Hui, Jeffrey, Bansal, Kshitij, Rao, Vikram, Ghiya, Rakesh, Wang, Emma, Ye, Ke, Sarr, Jean Michel, Preston, Melanie Moranski, Elish, Madeleine, Li, Steve, Kaku, Aakash, Gupta, Jigar, Pasupat, Ice, Juan, Da-Cheng, Someswar, Milan, M., Tejvi, Chen, Xinyun, Amini, Aida, Fabrikant, Alex, Chu, Eric, Dong, Xuanyi, Muthal, Amruta, Buthpitiya, Senaka, Jauhari, Sarthak, Khandelwal, Urvashi, Hitron, Ayal, Ren, Jie, Rinaldi, Larissa, Drath, Shahar, Dabush, Avigail, Jiang, Nan-Jiang, Godhia, Harshal, Sachs, Uli, Chen, Anthony, Fan, Yicheng, Taitelbaum, Hagai, Noga, Hila, Dai, Zhuyun, Wang, James, Hamer, Jenny, Ferng, Chun-Sung, Elkind, Chenel, Atias, Aviel, Lee, Paulina, Listík, Vít, Carlen, Mathias, van de Kerkhof, Jan, Pikus, Marcin, Zaher, Krunoslav, Müller, Paul, Zykova, Sasha, Stefanec, Richard, Gatsko, Vitaly, Hirnschall, Christoph, Sethi, Ashwin, Xu, Xingyu Federico, Ahuja, Chetan, Tsai, Beth, Stefanoiu, Anca, Feng, Bo, Dhandhania, Keshav, Katyal, Manish, Gupta, Akshay, Parulekar, Atharva, Pitta, Divya, Zhao, Jing, Bhatia, Vivaan, Bhavnani, Yashodha, Alhadlaq, Omar, Li, Xiaolin, Danenberg, Peter, Tu, Dennis, Pine, Alex, Filippova, Vera, Ghosh, Abhipso, Limonchik, Ben, Urala, Bhargava, Lanka, Chaitanya Krishna, Clive, Derik, Li, Edward, Wu, Hao, Hongtongsak, Kevin, Li, Ianna, Thakkar, Kalind, Omarov, Kuanysh, Majmundar, Kushal, Alverson, Michael, Kucharski, Michael, Patel, Mohak, Jain, Mudit, Zabelin, Maksim, Pelagatti, Paolo, Kohli, Rohan, Kumar, Saurabh, Kim, Joseph, Sankar, Swetha, Shah, Vineet, Ramachandruni, Lakshmi, Zeng, Xiangkai, Bariach, Ben, Weidinger, Laura, Vu, Tu, Andreev, Alek, He, Antoine, Hui, Kevin, Kashem, Sheleem, Subramanya, Amar, Hsiao, Sissie, Hassabis, Demis, Kavukcuoglu, Koray, Sadovsky, Adam, Le, Quoc, Strohman, Trevor, Wu, Yonghui, Petrov, Slav, Dean, Jeffrey, and Vinyals, Oriol
Subjects: Computer Science - Computation and Language, Computer Science - Artificial Intelligence, Computer Science - Computer Vision and Pattern Recognition
Abstract: This report introduces a new family of multimodal models, Gemini, that exhibit remarkable capabilities across image, audio, video, and text understanding. The Gemini family consists of Ultra, Pro, and Nano sizes, suitable for applications ranging from complex reasoning tasks to on-device memory-constrained use-cases. Evaluation on a broad range of benchmarks shows that our most-capable Gemini Ultra model advances the state of the art in 30 of 32 of these benchmarks - notably being the first model to achieve human-expert performance on the well-studied exam benchmark MMLU, and improving the state of the art in every one of the 20 multimodal benchmarks we examined. We believe that the new capabilities of the Gemini family in cross-modal reasoning and language understanding will enable a wide variety of use cases. We discuss our approach toward post-training and deploying Gemini models responsibly to users through services including Gemini, Gemini Advanced, Google AI Studio, and Cloud Vertex AI.
Published: 2023

4. Acceleration in Policy Optimization

Author: Chelu, Veronica, Zahavy, Tom, Guez, Arthur, Precup, Doina, and Flennerhag, Sebastian
Subjects: Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Statistics - Machine Learning
Abstract: We work towards a unifying paradigm for accelerating policy optimization methods in reinforcement learning (RL) by integrating foresight in the policy improvement step via optimistic and adaptive updates. Leveraging the connection between policy iteration and policy gradient methods, we view policy optimization algorithms as iteratively solving a sequence of surrogate objectives, local lower bounds on the original objective. We define optimism as predictive modelling of the future behavior of a policy, and adaptivity as taking immediate and anticipatory corrective actions to mitigate accumulating errors from overshooting predictions or delayed responses to change. We use this shared lens to jointly express other well-known algorithms, including model-based policy improvement based on forward search, and optimistic meta-learning algorithms. We analyze properties of this formulation, and show connections to other accelerated optimization algorithms. Then, we design an optimistic policy gradient algorithm, adaptive via meta-gradient learning, and empirically highlight several design choices pertaining to acceleration, in an illustrative task.
Published: 2023

5. Large-Scale Retrieval for Reinforcement Learning

Author: Humphreys, Peter C., Guez, Arthur, Tieleman, Olivier, Sifre, Laurent, Weber, Théophane, and Lillicrap, Timothy
Subjects: Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract: Effective decision making involves flexibly relating past experiences and relevant contextual information to a novel situation. In deep reinforcement learning (RL), the dominant paradigm is for an agent to amortise information that helps decision making into its network weights via gradient descent on training losses. Here, we pursue an alternative approach in which agents can utilise large-scale context sensitive database lookups to support their parametric computations. This allows agents to directly learn in an end-to-end manner to utilise relevant information to inform their outputs. In addition, new information can be attended to by the agent, without retraining, by simply augmenting the retrieval dataset. We study this approach for offline RL in 9x9 Go, a challenging game for which the vast combinatorial state space privileges generalisation over direct matching to past experiences. We leverage fast, approximate nearest neighbor techniques in order to retrieve relevant data from a set of tens of millions of expert demonstration states. Attending to this information provides a significant boost to prediction accuracy and game-play performance over simply using these demonstrations as training trajectories, providing a compelling demonstration of the value of large-scale retrieval in offline RL agents., Comment: Thirty-sixth Annual Conference on Neural Information Processing Systems (NeurIPS 2022), 16 pages
Published: 2022

6. COptiDICE: Offline Constrained Reinforcement Learning via Stationary Distribution Correction Estimation

Author: Lee, Jongmin, Paduraru, Cosmin, Mankowitz, Daniel J., Heess, Nicolas, Precup, Doina, Kim, Kee-Eung, and Guez, Arthur
Subjects: Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract: We consider the offline constrained reinforcement learning (RL) problem, in which the agent aims to compute a policy that maximizes expected return while satisfying given cost constraints, learning only from a pre-collected dataset. This problem setting is appealing in many real-world scenarios, where direct interaction with the environment is costly or risky, and where the resulting policy should comply with safety constraints. However, it is challenging to compute a policy that guarantees satisfying the cost constraints in the offline RL setting, since the off-policy evaluation inherently has an estimation error. In this paper, we present an offline constrained RL algorithm that optimizes the policy in the space of the stationary distribution. Our algorithm, COptiDICE, directly estimates the stationary distribution corrections of the optimal policy with respect to returns, while constraining the cost upper bound, with the goal of yielding a cost-conservative policy for actual constraint satisfaction. Experimental results show that COptiDICE attains better policies in terms of constraint satisfaction and return-maximization, outperforming baseline algorithms., Comment: 24 pages, 6 figures, Accepted at ICLR 2022 (spotlight)
Published: 2022

7. Retrieval-Augmented Reinforcement Learning

Author: Goyal, Anirudh, Friesen, Abram L., Banino, Andrea, Weber, Theophane, Ke, Nan Rosemary, Badia, Adria Puigdomenech, Guez, Arthur, Mirza, Mehdi, Humphreys, Peter C., Konyushkova, Ksenia, Sifre, Laurent, Valko, Michal, Osindero, Simon, Lillicrap, Timothy, Heess, Nicolas, and Blundell, Charles
Subjects: Computer Science - Machine Learning
Abstract: Most deep reinforcement learning (RL) algorithms distill experience into parametric behavior policies or value functions via gradient updates. While effective, this approach has several disadvantages: (1) it is computationally expensive, (2) it can take many updates to integrate experiences into the parametric model, (3) experiences that are not fully integrated do not appropriately influence the agent's behavior, and (4) behavior is limited by the capacity of the model. In this paper we explore an alternative paradigm in which we train a network to map a dataset of past experiences to optimal behavior. Specifically, we augment an RL agent with a retrieval process (parameterized as a neural network) that has direct access to a dataset of experiences. This dataset can come from the agent's past experiences, expert demonstrations, or any other relevant source. The retrieval process is trained to retrieve information from the dataset that may be useful in the current context, to help the agent achieve its goal faster and more efficiently. he proposed method facilitates learning agents that at test-time can condition their behavior on the entire dataset and not only the current state, or current trajectory. We integrate our method into two different RL agents: an offline DQN agent and an online R2D2 agent. In offline multi-task problems, we show that the retrieval-augmented DQN agent avoids task interference and learns faster than the baseline DQN agent. On Atari, we show that retrieval-augmented R2D2 learns significantly faster than the baseline R2D2 agent and achieves higher scores. We run extensive ablations to measure the contributions of the components of our proposed method.
Published: 2022

8. Muesli: Combining Improvements in Policy Optimization

Author: Hessel, Matteo, Danihelka, Ivo, Viola, Fabio, Guez, Arthur, Schmitt, Simon, Sifre, Laurent, Weber, Theophane, Silver, David, and van Hasselt, Hado
Subjects: Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract: We propose a novel policy update that combines regularized policy optimization with model learning as an auxiliary loss. The update (henceforth Muesli) matches MuZero's state-of-the-art performance on Atari. Notably, Muesli does so without using deep search: it acts directly with a policy network and has computation speed comparable to model-free baselines. The Atari results are complemented by extensive ablations, and by additional results on continuous control and 9x9 Go.
Published: 2021

9. Counterfactual Credit Assignment in Model-Free Reinforcement Learning

Author: Mesnard, Thomas, Weber, Théophane, Viola, Fabio, Thakoor, Shantanu, Saade, Alaa, Harutyunyan, Anna, Dabney, Will, Stepleton, Tom, Heess, Nicolas, Guez, Arthur, Moulines, Éric, Hutter, Marcus, Buesing, Lars, and Munos, Rémi
Subjects: Computer Science - Machine Learning
Abstract: Credit assignment in reinforcement learning is the problem of measuring an action's influence on future rewards. In particular, this requires separating skill from luck, i.e. disentangling the effect of an action on rewards from that of external factors and subsequent actions. To achieve this, we adapt the notion of counterfactuals from causality theory to a model-free RL setup. The key idea is to condition value functions on future events, by learning to extract relevant information from a trajectory. We formulate a family of policy gradient algorithms that use these future-conditional value functions as baselines or critics, and show that they are provably low variance. To avoid the potential bias from conditioning on future information, we constrain the hindsight information to not contain information about the agent's actions. We demonstrate the efficacy and validity of our algorithm on a number of illustrative and challenging problems.
Published: 2020

10. On the role of planning in model-based deep reinforcement learning

Author: Hamrick, Jessica B., Friesen, Abram L., Behbahani, Feryal, Guez, Arthur, Viola, Fabio, Witherspoon, Sims, Anthony, Thomas, Buesing, Lars, Veličković, Petar, and Weber, Théophane
Subjects: Computer Science - Artificial Intelligence, Computer Science - Machine Learning
Abstract: Model-based planning is often thought to be necessary for deep, careful reasoning and generalization in artificial agents. While recent successes of model-based reinforcement learning (MBRL) with deep function approximation have strengthened this hypothesis, the resulting diversity of model-based methods has also made it difficult to track which components drive success and why. In this paper, we seek to disentangle the contributions of recent methods by focusing on three questions: (1) How does planning benefit MBRL agents? (2) Within planning, what choices drive performance? (3) To what extent does planning improve generalization? To answer these questions, we study the performance of MuZero (Schrittwieser et al., 2019), a state-of-the-art MBRL algorithm with strong connections and overlapping components with many other MBRL algorithms. We perform a number of interventions and ablations of MuZero across a wide range of environments, including control tasks, Atari, and 9x9 Go. Our results suggest the following: (1) Planning is most useful in the learning process, both for policy updates and for providing a more useful data distribution. (2) Using shallow trees with simple Monte-Carlo rollouts is as performant as more complex methods, except in the most difficult reasoning tasks. (3) Planning alone is insufficient to drive strong generalization. These results indicate where and how to utilize planning in reinforcement learning settings, and highlight a number of open questions for future MBRL research., Comment: Published at ICLR 2021
Published: 2020

11. Beyond Tabula-Rasa: a Modular Reinforcement Learning Approach for Physically Embedded 3D Sokoban

Author: Karkus, Peter, Mirza, Mehdi, Guez, Arthur, Jaegle, Andrew, Lillicrap, Timothy, Buesing, Lars, Heess, Nicolas, and Weber, Theophane
Subjects: Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Robotics, Statistics - Machine Learning
Abstract: Intelligent robots need to achieve abstract objectives using concrete, spatiotemporally complex sensory information and motor control. Tabula rasa deep reinforcement learning (RL) has tackled demanding tasks in terms of either visual, abstract, or physical reasoning, but solving these jointly remains a formidable challenge. One recent, unsolved benchmark task that integrates these challenges is Mujoban, where a robot needs to arrange 3D warehouses generated from 2D Sokoban puzzles. We explore whether integrated tasks like Mujoban can be solved by composing RL modules together in a sense-plan-act hierarchy, where modules have well-defined roles similarly to classic robot architectures. Unlike classic architectures that are typically model-based, we use only model-free modules trained with RL or supervised learning. We find that our modular RL approach dramatically outperforms the state-of-the-art monolithic RL agent on Mujoban. Further, learned modules can be reused when, e.g., using a different robot platform to solve the same task. Together our results give strong evidence for the importance of research into modular RL designs. Project website: https://sites.google.com/view/modular-rl/
Published: 2020

12. Physically Embedded Planning Problems: New Challenges for Reinforcement Learning

Author: Mirza, Mehdi, Jaegle, Andrew, Hunt, Jonathan J., Guez, Arthur, Tunyasuvunakool, Saran, Muldal, Alistair, Weber, Théophane, Karkus, Peter, Racanière, Sébastien, Buesing, Lars, Lillicrap, Timothy, and Heess, Nicolas
Subjects: Computer Science - Artificial Intelligence, Computer Science - Machine Learning
Abstract: Recent work in deep reinforcement learning (RL) has produced algorithms capable of mastering challenging games such as Go, chess, or shogi. In these works the RL agent directly observes the natural state of the game and controls that state directly with its actions. However, when humans play such games, they do not just reason about the moves but also interact with their physical environment. They understand the state of the game by looking at the physical board in front of them and modify it by manipulating pieces using touch and fine-grained motor control. Mastering complicated physical systems with abstract goals is a central challenge for artificial intelligence, but it remains out of reach for existing RL algorithms. To encourage progress towards this goal we introduce a set of physically embedded planning problems and make them publicly available. We embed challenging symbolic tasks (Sokoban, tic-tac-toe, and Go) in a physics engine to produce a set of tasks that require perception, reasoning, and motor control over long time horizons. Although existing RL algorithms can tackle the symbolic versions of these tasks, we find that they struggle to master even the simplest of their physically embedded counterparts. As a first step towards characterizing the space of solution to these tasks, we introduce a strong baseline that uses a pre-trained expert game player to provide hints in the abstract space to an RL agent's policy while training it on the full sensorimotor control task. The resulting agent solves many of the tasks, underlining the need for methods that bridge the gap between abstract planning and embodied control. See illustrating video at https://youtu.be/RwHiHlym_1k., Comment: 17 pages + appendix. Updated text and references
Published: 2020

13. Value-driven Hindsight Modelling

Author: Guez, Arthur, Viola, Fabio, Weber, Théophane, Buesing, Lars, Kapturowski, Steven, Precup, Doina, Silver, David, and Heess, Nicolas
Subjects: Computer Science - Machine Learning, Statistics - Machine Learning
Abstract: Value estimation is a critical component of the reinforcement learning (RL) paradigm. The question of how to effectively learn value predictors from data is one of the major problems studied by the RL community, and different approaches exploit structure in the problem domain in different ways. Model learning can make use of the rich transition structure present in sequences of observations, but this approach is usually not sensitive to the reward function. In contrast, model-free methods directly leverage the quantity of interest from the future, but receive a potentially weak scalar signal (an estimate of the return). We develop an approach for representation learning in RL that sits in between these two extremes: we propose to learn what to model in a way that can directly help value prediction. To this end, we determine which features of the future trajectory provide useful information to predict the associated return. This provides tractable prediction targets that are directly relevant for a task, and can thus accelerate learning the value function. The idea can be understood as reasoning, in hindsight, about which aspects of the future observations could help past value prediction. We show how this can help dramatically even in simple policy evaluation settings. We then test our approach at scale in challenging domains, including on 57 Atari 2600 games., Comment: 9 pages + reference + appendix. NeurIPS 2020 version
Published: 2020

14. Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model

Author: Schrittwieser, Julian, Antonoglou, Ioannis, Hubert, Thomas, Simonyan, Karen, Sifre, Laurent, Schmitt, Simon, Guez, Arthur, Lockhart, Edward, Hassabis, Demis, Graepel, Thore, Lillicrap, Timothy, and Silver, David
Subjects: Computer Science - Machine Learning, Statistics - Machine Learning
Abstract: Constructing agents with planning capabilities has long been one of the main challenges in the pursuit of artificial intelligence. Tree-based planning methods have enjoyed huge success in challenging domains, such as chess and Go, where a perfect simulator is available. However, in real-world problems the dynamics governing the environment are often complex and unknown. In this work we present the MuZero algorithm which, by combining a tree-based search with a learned model, achieves superhuman performance in a range of challenging and visually complex domains, without any knowledge of their underlying dynamics. MuZero learns a model that, when applied iteratively, predicts the quantities most directly relevant to planning: the reward, the action-selection policy, and the value function. When evaluated on 57 different Atari games - the canonical video game environment for testing AI techniques, in which model-based planning approaches have historically struggled - our new algorithm achieved a new state of the art. When evaluated on Go, chess and shogi, without any knowledge of the game rules, MuZero matched the superhuman performance of the AlphaZero algorithm that was supplied with the game rules.
Published: 2019
Full Text: View/download PDF

15. Augmenting learning using symmetry in a biologically-inspired domain

Author: Mishra, Shruti, Abdolmaleki, Abbas, Guez, Arthur, Trochim, Piotr, and Precup, Doina
Subjects: Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Computer Science - Robotics, Statistics - Machine Learning
Abstract: Invariances to translation, rotation and other spatial transformations are a hallmark of the laws of motion, and have widespread use in the natural sciences to reduce the dimensionality of systems of equations. In supervised learning, such as in image classification tasks, rotation, translation and scale invariances are used to augment training datasets. In this work, we use data augmentation in a similar way, exploiting symmetry in the quadruped domain of the DeepMind control suite (Tassa et al. 2018) to add to the trajectories experienced by the actor in the actor-critic algorithm of Abdolmaleki et al. (2018). In a data-limited regime, the agent using a set of experiences augmented through symmetry is able to learn faster. Our approach can be used to inject knowledge of invariances in the domain and task to augment learning in robots, and more generally, to speed up learning in realistic robotics applications.
Published: 2019

16. An investigation of model-free planning

Author: Guez, Arthur, Mirza, Mehdi, Gregor, Karol, Kabra, Rishabh, Racanière, Sébastien, Weber, Théophane, Raposo, David, Santoro, Adam, Orseau, Laurent, Eccles, Tom, Wayne, Greg, Silver, David, and Lillicrap, Timothy
Subjects: Computer Science - Machine Learning, Computer Science - Artificial Intelligence, Statistics - Machine Learning
Abstract: The field of reinforcement learning (RL) is facing increasingly challenging domains with combinatorial complexity. For an RL agent to address these challenges, it is essential that it can plan effectively. Prior work has typically utilized an explicit model of the environment, combined with a specific planning algorithm (such as tree search). More recently, a new family of methods have been proposed that learn how to plan, by providing the structure for planning via an inductive bias in the function approximator (such as a tree structured neural network), trained end-to-end by a model-free RL algorithm. In this paper, we go even further, and demonstrate empirically that an entirely model-free approach, without special structure beyond standard neural network components such as convolutional networks and LSTMs, can learn to exhibit many of the characteristics typically associated with a model-based planner. We measure our agent's effectiveness at planning in terms of its ability to generalize across a combinatorial and irreversible state space, its data efficiency, and its ability to utilize additional thinking time. We find that our agent has many of the characteristics that one might expect to find in a planning algorithm. Furthermore, it exceeds the state-of-the-art in challenging combinatorial domains such as Sokoban and outperforms other model-free approaches that utilize strong inductive biases toward planning.
Published: 2019

17. Woulda, Coulda, Shoulda: Counterfactually-Guided Policy Search

Author: Buesing, Lars, Weber, Theophane, Zwols, Yori, Racaniere, Sebastien, Guez, Arthur, Lespiau, Jean-Baptiste, and Heess, Nicolas
Subjects: Computer Science - Machine Learning, Statistics - Machine Learning
Abstract: Learning policies on data synthesized by models can in principle quench the thirst of reinforcement learning algorithms for large amounts of real experience, which is often costly to acquire. However, simulating plausible experience de novo is a hard problem for many complex environments, often resulting in biases for model-based policy evaluation and search. Instead of de novo synthesis of data, here we assume logged, real experience and model alternative outcomes of this experience under counterfactual actions, actions that were not actually taken. Based on this, we propose the Counterfactually-Guided Policy Search (CF-GPS) algorithm for learning policies in POMDPs from off-policy experience. It leverages structural causal models for counterfactual evaluation of arbitrary policies on individual off-policy episodes. CF-GPS can improve on vanilla model-based RL algorithms by making use of available logged data to de-bias model predictions. In contrast to off-policy algorithms based on Importance Sampling which re-weight data, CF-GPS leverages a model to explicitly consider alternative outcomes, allowing the algorithm to make better use of experience data. We find empirically that these advantages translate into improved policy evaluation and search results on a non-trivial grid-world task. Finally, we show that CF-GPS generalizes the previously proposed Guided Policy Search and that reparameterization-based algorithms such Stochastic Value Gradient can be interpreted as counterfactual methods.
Published: 2018

18. Learning to Search with MCTSnets

Author: Guez, Arthur, Weber, Théophane, Antonoglou, Ioannis, Simonyan, Karen, Vinyals, Oriol, Wierstra, Daan, Munos, Rémi, and Silver, David
Subjects: Computer Science - Artificial Intelligence, Computer Science - Machine Learning, Statistics - Machine Learning
Abstract: Planning problems are among the most important and well-studied problems in artificial intelligence. They are most typically solved by tree search algorithms that simulate ahead into the future, evaluate future states, and back-up those evaluations to the root of a search tree. Among these algorithms, Monte-Carlo tree search (MCTS) is one of the most general, powerful and widely used. A typical implementation of MCTS uses cleverly designed rules, optimized to the particular characteristics of the domain. These rules control where the simulation traverses, what to evaluate in the states that are reached, and how to back-up those evaluations. In this paper we instead learn where, what and how to search. Our architecture, which we call an MCTSnet, incorporates simulation-based search inside a neural network, by expanding, evaluating and backing-up a vector embedding. The parameters of the network are trained end-to-end using gradient-based optimisation. When applied to small searches in the well known planning problem Sokoban, the learned search algorithm significantly outperformed MCTS baselines., Comment: ICML 2018 (camera-ready version)
Published: 2018

19. Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithm

Author: Silver, David, Hubert, Thomas, Schrittwieser, Julian, Antonoglou, Ioannis, Lai, Matthew, Guez, Arthur, Lanctot, Marc, Sifre, Laurent, Kumaran, Dharshan, Graepel, Thore, Lillicrap, Timothy, Simonyan, Karen, and Hassabis, Demis
Subjects: Computer Science - Artificial Intelligence, Computer Science - Learning
Abstract: The game of chess is the most widely-studied domain in the history of artificial intelligence. The strongest programs are based on a combination of sophisticated search techniques, domain-specific adaptations, and handcrafted evaluation functions that have been refined by human experts over several decades. In contrast, the AlphaGo Zero program recently achieved superhuman performance in the game of Go, by tabula rasa reinforcement learning from games of self-play. In this paper, we generalise this approach into a single AlphaZero algorithm that can achieve, tabula rasa, superhuman performance in many challenging domains. Starting from random play, and given no domain knowledge except the game rules, AlphaZero achieved within 24 hours a superhuman level of play in the games of chess and shogi (Japanese chess) as well as Go, and convincingly defeated a world-champion program in each case.
Published: 2017

20. Imagination-Augmented Agents for Deep Reinforcement Learning

Author: Weber, Théophane, Racanière, Sébastien, Reichert, David P., Buesing, Lars, Guez, Arthur, Rezende, Danilo Jimenez, Badia, Adria Puigdomènech, Vinyals, Oriol, Heess, Nicolas, Li, Yujia, Pascanu, Razvan, Battaglia, Peter, Hassabis, Demis, Silver, David, and Wierstra, Daan
Subjects: Computer Science - Learning, Computer Science - Artificial Intelligence, Statistics - Machine Learning
Abstract: We introduce Imagination-Augmented Agents (I2As), a novel architecture for deep reinforcement learning combining model-free and model-based aspects. In contrast to most existing model-based reinforcement learning and planning methods, which prescribe how a model should be used to arrive at a policy, I2As learn to interpret predictions from a learned environment model to construct implicit plans in arbitrary ways, by using the predictions as additional context in deep policy networks. I2As show improved data efficiency, performance, and robustness to model misspecification compared to several baselines.
Published: 2017

21. The Predictron: End-To-End Learning and Planning

Author: Silver, David, van Hasselt, Hado, Hessel, Matteo, Schaul, Tom, Guez, Arthur, Harley, Tim, Dulac-Arnold, Gabriel, Reichert, David, Rabinowitz, Neil, Barreto, Andre, and Degris, Thomas
Subjects: Computer Science - Learning, Computer Science - Artificial Intelligence, Computer Science - Neural and Evolutionary Computing
Abstract: One of the key challenges of artificial intelligence is to learn models that are effective in the context of planning. In this document we introduce the predictron architecture. The predictron consists of a fully abstract model, represented by a Markov reward process, that can be rolled forward multiple "imagined" planning steps. Each forward pass of the predictron accumulates internal rewards and values over multiple planning depths. The predictron is trained end-to-end so as to make these accumulated values accurately approximate the true value function. We applied the predictron to procedurally generated random mazes and a simulator for the game of pool. The predictron yielded significantly more accurate predictions than conventional deep neural network architectures., Comment: Camera-ready version, ICML 2017, with supplement
Published: 2016

22. Learning values across many orders of magnitude

Author: van Hasselt, Hado, Guez, Arthur, Hessel, Matteo, Mnih, Volodymyr, and Silver, David
Subjects: Computer Science - Learning, Computer Science - Artificial Intelligence, Computer Science - Neural and Evolutionary Computing, Statistics - Machine Learning
Abstract: Most learning algorithms are not invariant to the scale of the function that is being approximated. We propose to adaptively normalize the targets used in learning. This is useful in value-based reinforcement learning, where the magnitude of appropriate value approximations can change over time when we update the policy of behavior. Our main motivation is prior work on learning to play Atari games, where the rewards were all clipped to a predetermined range. This clipping facilitates learning across many different games with a single learning algorithm, but a clipped reward function can result in qualitatively different behavior. Using the adaptive normalization we can remove this domain-specific heuristic without diminishing overall performance., Comment: Paper accepted for publication at NIPS 2016. This version includes the appendix
Published: 2016

23. Adaptive planning in human search

Author: Krusche, Moritz J.F., Schulz, Eric, Guez, Arthur, and Speekenbrink, Maarten
Subjects: Planning, Decision Making, Tree Search, Forag-ing, Reinforcement Learning, Monte Carlo Sampling
Abstract: How do people plan ahead when searching for rewards? Weinvestigate planning in a foraging task in which participantssearch for rewards on an infinite two-dimensional grid. Ourresults show that their search is best-described by a modelwhich searches at least 3 steps ahead. Furthermore, partici-pants do not seem to update their beliefs during planning, butrather treat their initial beliefs as given, a strategy similar to aheuristic called root-sampling. This planning algorithm corre-sponds well with participants’ behavior in test problems withrestricted movement and varying degrees of information, out-performing more complex models. These results enrich ourunderstanding of adaptive planning in complex environments
Published: 2018

24. Increasing the Action Gap: New Operators for Reinforcement Learning

Author: Bellemare, Marc G., Ostrovski, Georg, Guez, Arthur, Thomas, Philip S., and Munos, Rémi
Subjects: Computer Science - Artificial Intelligence, Computer Science - Learning
Abstract: This paper introduces new optimality-preserving operators on Q-functions. We first describe an operator for tabular representations, the consistent Bellman operator, which incorporates a notion of local policy consistency. We show that this local consistency leads to an increase in the action gap at each state; increasing this gap, we argue, mitigates the undesirable effects of approximation and estimation errors on the induced greedy policies. This operator can also be applied to discretized continuous space and time problems, and we provide empirical results evidencing superior performance in this context. Extending the idea of a locally consistent operator, we then derive sufficient conditions for an operator to preserve optimality, leading to a family of operators which includes our consistent Bellman operator. As corollaries we provide a proof of optimality for Baird's advantage learning algorithm and derive other gap-increasing operators with interesting properties. We conclude with an empirical study on 60 Atari 2600 games illustrating the strong potential of these new operators.
Published: 2015

25. Deep Reinforcement Learning with Double Q-learning

Author: van Hasselt, Hado, Guez, Arthur, and Silver, David
Subjects: Computer Science - Learning
Abstract: The popular Q-learning algorithm is known to overestimate action values under certain conditions. It was not previously known whether, in practice, such overestimations are common, whether they harm performance, and whether they can generally be prevented. In this paper, we answer all these questions affirmatively. In particular, we first show that the recent DQN algorithm, which combines Q-learning with a deep neural network, suffers from substantial overestimations in some games in the Atari 2600 domain. We then show that the idea behind the Double Q-learning algorithm, which was introduced in a tabular setting, can be generalized to work with large-scale function approximation. We propose a specific adaptation to the DQN algorithm and show that the resulting algorithm not only reduces the observed overestimations, as hypothesized, but that this also leads to much better performance on several games., Comment: AAAI 2016
Published: 2015

26. Better Optimism By Bayes: Adaptive Planning with Rich Models

Author: Guez, Arthur, Silver, David, and Dayan, Peter
Subjects: Computer Science - Artificial Intelligence, Computer Science - Learning, Statistics - Machine Learning
Abstract: The computational costs of inference and planning have confined Bayesian model-based reinforcement learning to one of two dismal fates: powerful Bayes-adaptive planning but only for simplistic models, or powerful, Bayesian non-parametric models but using simple, myopic planning strategies such as Thompson sampling. We ask whether it is feasible and truly beneficial to combine rich probabilistic models with a closer approximation to fully Bayesian planning. First, we use a collection of counterexamples to show formal problems with the over-optimism inherent in Thompson sampling. Then we leverage state-of-the-art techniques in efficient Bayes-adaptive planning and non-parametric Bayesian methods to perform qualitatively better than both existing conventional algorithms and Thompson sampling on two contextual bandit-like problems., Comment: 11 pages, 11 figures
Published: 2014

27. Mastering Atari, Go, chess and shogi by planning with a learned model

Author: Schrittwieser, Julian, Antonoglou, Ioannis, Hubert, Thomas, Simonyan, Karen, Sifre, Laurent, Schmitt, Simon, and Guez, Arthur
Subjects: Artificial intelligence -- Research, Binary searching -- Usage, Games -- Technology application, Reinforcement learning (Machine learning) -- Models, Artificial intelligence, Technology application, Environmental issues, Science and technology, Zoology and wildlife conservation
Abstract: Constructing agents with planning capabilities has long been one of the main challenges in the pursuit of artificial intelligence. Tree-based planning methods have enjoyed huge success in challenging domains, such as chess.sup.1 and Go.sup.2, where a perfect simulator is available. However, in real-world problems, the dynamics governing the environment are often complex and unknown. Here we present the MuZero algorithm, which, by combining a tree-based search with a learned model, achieves superhuman performance in a range of challenging and visually complex domains, without any knowledge of their underlying dynamics. The MuZero algorithm learns an iterable model that produces predictions relevant to planning: the action-selection policy, the value function and the reward. When evaluated on 57 different Atari games.sup.3--the canonical video game environment for testing artificial intelligence techniques, in which model-based planning approaches have historically struggled.sup.4--the MuZero algorithm achieved state-of-the-art performance. When evaluated on Go, chess and shogi--canonical environments for high-performance planning--the MuZero algorithm matched, without any knowledge of the game dynamics, the superhuman performance of the AlphaZero algorithm.sup.5 that was supplied with the rules of the game. A reinforcement-learning algorithm that combines a tree-based search with a learned model achieves superhuman performance in high-performance planning and visually complex domains, without any knowledge of their underlying dynamics., Author(s): Julian Schrittwieser [sup.1] , Ioannis Antonoglou [sup.1] [sup.2] , Thomas Hubert [sup.1] , Karen Simonyan [sup.1] , Laurent Sifre [sup.1] , Simon Schmitt [sup.1] , Arthur Guez [sup.1] , [...]
Published: 2020
Full Text: View/download PDF

28. Efficient Bayes-Adaptive Reinforcement Learning using Sample-Based Search

Author: Guez, Arthur, Silver, David, and Dayan, Peter
Subjects: Computer Science - Learning, Computer Science - Artificial Intelligence, Statistics - Machine Learning
Abstract: Bayesian model-based reinforcement learning is a formally elegant approach to learning optimal behaviour under model uncertainty, trading off exploration and exploitation in an ideal way. Unfortunately, finding the resulting Bayes-optimal policies is notoriously taxing, since the search space becomes enormous. In this paper we introduce a tractable, sample-based method for approximate Bayes-optimal planning which exploits Monte-Carlo tree search. Our approach outperformed prior Bayesian model-based RL algorithms by a significant margin on several well-known benchmark problems -- because it avoids expensive applications of Bayes rule within the search tree by lazily sampling models from the current beliefs. We illustrate the advantages of our approach by showing it working in an infinite state space domain which is qualitatively out of reach of almost all previous work in Bayesian exploration., Comment: 14 pages, 7 figures, includes supplementary material. Advances in Neural Information Processing Systems (NIPS) 2012
Published: 2012

29. A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play

Author: Silver, David, Hubert, Thomas, Schrittwieser, Julian, Antonogloux, Ioannis, Lai, Matthew, Guez, Arthur, Lanctot, Marc, Sifre, Laurent, Kumaran, Dharshan, Graepel, Thore, Lillicrap, Timothy, Simonyan, Karen, and Hassabis, Demis
Published: 2018

30. Mastering the game of Go without human knowledge

Author: Silver, David, Schrittwieser, Julian, Simonyan, Karen, Antonoglou, Ioannis, Huang, Aja, Guez, Arthur, Hubert, Thomas, Baker, Lucas, Lai, Matthew, Bolton, Adrian, Chen, Yutian, Lillicrap, Timothy, Hui, Fan, Sifre, Laurent, van den Driessche, George, Graepel, Thore, and Hassabis, Demis
Subjects: Artificial intelligence -- Forecasts and trends, Go (Game) -- Technology application, Computer games -- Forecasts and trends, Computer game, Market trend/market analysis, Artificial intelligence, Technology application, Environmental issues, Science and technology, Zoology and wildlife conservation, AlphaGo (Computer game) -- Forecasts and trends
Abstract: A long-standing goal of artificial intelligence is an algorithm that learns, tabula rasa, superhuman proficiency in challenging domains. Recently, AlphaGo became the first program to defeat a world champion in the game of Go. The tree search in AlphaGo evaluated positions and selected moves using deep neural networks. These neural networks were trained by supervised learning from human expert moves, and by reinforcement learning from self-play. Here we introduce an algorithm based solely on reinforcement learning, without human data, guidance or domain knowledge beyond game rules. AlphaGo becomes its own teacher: a neural network is trained to predict AlphaGos own move selections and also the winner of AlphaGos games. This neural network improves the strength of the tree search, resulting in higher quality move selection and stronger self-play in the next iteration. Starting tabula rasa:, our new program AlphaGo Zero achieved superhuman performance, winning 1000 against the previously published, champion-defeating AlphaGo., Author(s): David Silver (corresponding author) [1]; Julian Schrittwieser [1]; Karen Simonyan [1]; Ioannis Antonoglou [1]; Aja Huang [1]; Arthur Guez [1]; Thomas Hubert [1]; Lucas Baker [1]; Matthew Lai [1]; [...]
Published: 2017
Full Text: View/download PDF

31. Optimism and Adaptivity in Policy Optimization

Author: Chelu, Veronica, Zahavy, Tom, Guez, Arthur, Precup, Doina, and Flennerhag, Sebastian
Subjects: FOS: Computer and information sciences, Computer Science - Machine Learning, Artificial Intelligence (cs.AI), Computer Science - Artificial Intelligence, Statistics - Machine Learning, Machine Learning (stat.ML), Machine Learning (cs.LG)
Abstract: We work towards a unifying paradigm for accelerating policy optimization methods in reinforcement learning (RL) through \emph{optimism} \& \emph{adaptivity}. Leveraging the deep connection between policy iteration and policy gradient methods, we recast seemingly unrelated policy optimization algorithms as the repeated application of two interleaving steps (i) an \emph{optimistic policy improvement operator} maps a prior policy $π_t$ to a hypothesis $π_{t+1}$ using a \emph{gradient ascent prediction}, followed by (ii) a \emph{hindsight adaptation} of the optimistic prediction based on a partial evaluation of the performance of $π_{t+1}$. We use this shared lens to jointly express other well-known algorithms, including soft and optimistic policy iteration, natural actor-critic methods, model-based policy improvement based on forward search, and meta-learning algorithms. By doing so, we shed light on collective theoretical properties related to acceleration via optimism \& adaptivity. Building on these insights, we design an \emph{adaptive \& optimistic policy gradient} algorithm via meta-gradient learning, and empirically highlight several design choices pertaining to optimism, in an illustrative task.
Published: 2023
Full Text: View/download PDF

32. Adaptive control of epileptiform excitability in an in vitro model of limbic seizures

Author: Panuccio, Gabriella, Guez, Arthur, Vincent, Robert, Avoli, Massimo, and Pineau, Joelle
Published: 2013
Full Text: View/download PDF

33. Mastering the game of Go with deep neural networks and tree search

Author: Silver, David, Huang, Aja, Maddison, Chris J., Guez, Arthur, Sifre, Laurent, van den Driessche, George, Schrittwieser, Julian, Antonoglou, Ionannis, Panneershelvam, Veda, Lanctot, Marc, Dieleman, Sander, Grewe, Dominik, Nham, John, Kalchbrenner, Nal, Sutskever, Ilya, Lillicrap, Timothy, and Leach, Madeleine
Subjects: Go (Game) -- Technology application, Computer games -- Product development, Neural networks -- Product development, Neural network, Computer game, Technology application, Environmental issues, Science and technology, Zoology and wildlife conservation, AlphaGo (Computer game) -- Product development
Abstract: The game of Go has long been viewed as the most challenging of classic games for artificial intelligence owing to its enormous search space and the difficulty of evaluating board positions and moves. Here we introduce a new approach to computer Go that uses 'value networks' to evaluate board positions and 'policy networks' to select moves. These deep neural networks are trained by a novel combination of supervised learning from human expert games, and reinforcement learning from games of self-play. Without any lookahead search, the neural networks play Go at the level of state-of-the-art Monte Carlo tree search programs that simulate thousands of random games of self-play. We also introduce a new search algorithm that combines Monte Carlo simulation with value and policy networks. Using this search algorithm, our program AlphaGo achieved a 99.8% winning rate against other Go programs, and defeated the human European Go champion by 5 games to 0. This is the first time that a computer program has defeated a human professional player in the full-sized game of Go, a feat previously thought to be at least a decade away. A computer Go program based on deep neural networks defeats a human professional player to achieve one of the grand challenges of artificial intelligence. AlphaGo computer beats Go champion The victory in 1997 of the chess-playing computer Deep Blue in a six-game series against the then world champion Gary Kasparov was seen as a significant milestone in the development of artificial intelligence. An even greater challenge remained -- the ancient game of Go. Despite decades of refinement, until recently the strongest computers were still playing Go at the level of human amateurs. Enter AlphaGo. Developed by Google DeepMind, this program uses deep neural networks to mimic expert players, and further improves its performance by learning from games played against itself. AlphaGo has achieved a 99% win rate against the strongest other Go programs, and defeated the reigning European champion Fan Hui 5-0 in a tournament match. This is the first time that a computer program has defeated a human professional player in even games, on a full, 19 x 19 board, in even games with no handicap., Author(s): David Silver [sup.1] , Aja Huang [sup.1] , Chris J. Maddison [sup.1] , Arthur Guez [sup.1] , Laurent Sifre [sup.1] , George van den Driessche [sup.1] , Julian Schrittwieser [...]
Published: 2016
Full Text: View/download PDF

34. Adaptive planning in human search

Author: Krusche, Moritz J. F., primary, Schulz, Eric, additional, Guez, Arthur, additional, and Speekenbrink, Maarten, additional
Published: 2018
Full Text: View/download PDF

35. Deep Reinforcement Learning with Double Q-Learning

Author: van Hasselt, Hado, Guez, Arthur, and Silver, David
Subjects: FOS: Computer and information sciences, Computer Science - Learning, General Medicine, Machine Learning (cs.LG)
Abstract: The popular Q-learning algorithm is known to overestimate action values under certain conditions. It was not previously known whether, in practice, such overestimations are common, whether they harm performance, and whether they can generally be prevented. In this paper, we answer all these questions affirmatively. In particular, we first show that the recent DQN algorithm, which combines Q-learning with a deep neural network, suffers from substantial overestimations in some games in the Atari 2600 domain. We then show that the idea behind the Double Q-learning algorithm, which was introduced in a tabular setting, can be generalized to work with large-scale function approximation. We propose a specific adaptation to the DQN algorithm and show that the resulting algorithm not only reduces the observed overestimations, as hypothesized, but that this also leads to much better performance on several games., Comment: AAAI 2016
Published: 2016
Full Text: View/download PDF

36. Increasing the Action Gap: New Operators for Reinforcement Learning

Author: Bellemare, Marc G., Ostrovski, Georg, Guez, Arthur, Thomas, Philip S., and Munos, R��mi
Subjects: FOS: Computer and information sciences, Computer Science - Learning, Artificial Intelligence (cs.AI), Computer Science - Artificial Intelligence, General Medicine, Machine Learning (cs.LG)
Abstract: This paper introduces new optimality-preserving operators on Q-functions. We first describe an operator for tabular representations, the consistent Bellman operator, which incorporates a notion of local policy consistency. We show that this local consistency leads to an increase in the action gap at each state; increasing this gap, we argue, mitigates the undesirable effects of approximation and estimation errors on the induced greedy policies. This operator can also be applied to discretized continuous space and time problems, and we provide empirical results evidencing superior performance in this context. Extending the idea of a locally consistent operator, we then derive sufficient conditions for an operator to preserve optimality, leading to a family of operators which includes our consistent Bellman operator. As corollaries we provide a proof of optimality for Baird's advantage learning algorithm and derive other gap-increasing operators with interesting properties. We conclude with an empirical study on 60 Atari 2600 games illustrating the strong potential of these new operators.
Published: 2016
Full Text: View/download PDF

37. Increasing the Action Gap: New Operators for Reinforcement Learning

Author: G. Bellemare, Marc, primary, Ostrovski, Georg, additional, Guez, Arthur, additional, Thomas, Philip, additional, and Munos, Remi, additional
Published: 2016
Full Text: View/download PDF

38. Multi-tasking SLAM

Author: Guez, Arthur, primary and Pineau, Joelle, additional
Published: 2010
Full Text: View/download PDF

39. TREATING EPILEPSY VIA ADAPTIVE NEUROSTIMULATION: A REINFORCEMENT LEARNING APPROACH

Author: PINEAU, JOELLE, primary, GUEZ, ARTHUR, additional, VINCENT, ROBERT, additional, PANUCCIO, GABRIELLA, additional, and AVOLI, MASSIMO, additional
Published: 2009
Full Text: View/download PDF

Catalog

Books, media, physical & digital resources

See catalog results

Searchworks

Select search scope, currently: Articles Catalog books, media & more in Jio Institute collections Articles journal articles & other e-resources

Search

Search Constraints

Refine your results

Search Limiters

Topic

Publication Year Range

Language

Publication Type

Journal

Database

Publisher

39 results on '"Guez, Arthur"'

Search Results

Catalog

Select search scope, currently: Articles

Catalog

books, media & more in Jio Institute collections

Articles

journal articles & other e-resources