Search

Your search keyword '"Smith, Eric"' showing total 12,084 results
Start Over Author "Smith, Eric"

Refine your results

more »

more »

more »

more »

more »

more »
12,084 results on '"Smith, Eric"'

105. Reproductive inequality in humans and other mammals

Author

Ross, Cody T, Hooper, Paul L, Smith, Jennifer E, Jaeggi, Adrian V, Smith, Eric Alden, Gavrilets, Sergey, Zohora, Fatema tuz, Ziker, John, Xygalatas, Dimitris, Wroblewski, Emily E, Wood, Brian, Winterhalder, Bruce, Willführ, Kai P, Willard, Aiyana K, Walker, Kara, von Rueden, Christopher, Voland, Eckart, Valeggia, Claudia, Vaitla, Bapu, Urlacher, Samuel, Towner, Mary, Sum, Chun-Yi, Sugiyama, Lawrence S, Strier, Karen B, Starkweather, Kathrine, Major-Smith, Daniel, Shenk, Mary, Sear, Rebecca, Seabright, Edmond, Schacht, Ryan, Scelza, Brooke, Scaggs, Shane, Salerno, Jonathan, Revilla-Minaya, Caissa, Redhead, Daniel, Pusey, Anne, Purzycki, Benjamin Grant, Power, Eleanor A, Pisor, Anne, Pettay, Jenni, Perry, Susan, Page, Abigail E, Pacheco-Cobos, Luis, Oths, Kathryn, Oh, Seung-Yun, Nolin, David, Nettle, Daniel, Moya, Cristina, Migliano, Andrea Bamberg, Mertens, Karl J, McNamara, Rita A, McElreath, Richard, Mattison, Siobhan, Massengill, Eric, Marlowe, Frank, Madimenos, Felicia, Macfarlan, Shane, Lummaa, Virpi, Lizarralde, Roberto, Liu, Ruizhe, Liebert, Melissa A, Lew-Levy, Sheina, Leslie, Paul, Lanning, Joseph, Kramer, Karen, Koster, Jeremy, Kaplan, Hillard S, Jamsranjav, Bayarsaikhan, Hurtado, A Magdalena, Hill, Kim, Hewlett, Barry, Helle, Samuli, Headland, Thomas, Headland, Janet, Gurven, Michael, Grimalda, Gianluca, Greaves, Russell, Golden, Christopher D, Godoy, Irene, Gibson, Mhairi, Mouden, Claire El, Dyble, Mark, Draper, Patricia, Downey, Sean, DeMarco, Angelina L, Davis, Helen Elizabeth, Crabtree, Stefani, Cortez, Carmen, Colleran, Heidi, Cohen, Emma, Clark, Gregory, Clark, Julia, Caudell, Mark A, Carminito, Chelsea E, Bunce, John, Boyette, Adam, Bowles, Samuel, Blumenfield, Tami, Beheim, Bret, and Beckerman, Stephen

Subjects
Contraception/Reproduction, Reproductive health and childbirth, Reduced Inequalities, Animals, Humans, Female, Male, Reproduction, Sex Characteristics, Marriage, Mammals, Sexual Behavior, Animal, egalitarian syndrome, inequality, mating systems, monogamy, reproductive skew
Abstract

To address claims of human exceptionalism, we determine where humans fit within the greater mammalian distribution of reproductive inequality. We show that humans exhibit lower reproductive skew (i.e., inequality in the number of surviving offspring) among males and smaller sex differences in reproductive skew than most other mammals, while nevertheless falling within the mammalian range. Additionally, female reproductive skew is higher in polygynous human populations than in polygynous nonhumans mammals on average. This patterning of skew can be attributed in part to the prevalence of monogamy in humans compared to the predominance of polygyny in nonhuman mammals, to the limited degree of polygyny in the human societies that practice it, and to the importance of unequally held rival resources to women's fitness. The muted reproductive inequality observed in humans appears to be linked to several unusual characteristics of our species-including high levels of cooperation among males, high dependence on unequally held rival resources, complementarities between maternal and paternal investment, as well as social and legal institutions that enforce monogamous norms.

Published
2023

106. Action Functional Gradient Descent algorithm for estimating escape paths in Stochastic Chemical Reaction Networks

Author

Gagrani, Praful and Smith, Eric

Subjects
Condensed Matter - Statistical Mechanics, Mathematical Physics
Abstract

We first derive the Hamilton-Jacobi theory underlying continuous-time Markov processes, and then use the construction to develop a variational algorithm for estimating escape (least improbable or first passage) paths for a generic stochastic chemical reaction network that exhibits multiple fixed points. The design of our algorithm is such that it is independent of the underlying dimensionality of the system, the discretization control parameters are updated towards the continuum limit, and there is an easy-to-calculate measure for the correctness of its solution. We consider several applications of the algorithm and verify them against computationally expensive means such as the shooting method and stochastic simulation. While we employ theoretical techniques from mathematical physics, numerical optimization and chemical reaction network theory, we hope that our work finds practical applications with an inter-disciplinary audience including chemists, biologists, optimal control theorists and game theorists., Comment: 40 pages, 21 figures

Published
2022
Full Text
View/download PDF

107. Toxicity in Multilingual Machine Translation at Scale

Author

Costa-jussà, Marta R., Smith, Eric, Ropers, Christophe, Licht, Daniel, Maillard, Jean, Ferrando, Javier, and Escolano, Carlos

Subjects
Computer Science - Computation and Language, I.2.7
Abstract

Machine Translation systems can produce different types of errors, some of which are characterized as critical or catastrophic due to the specific negative impact that they can have on users. In this paper we focus on one type of critical error: added toxicity. We evaluate and analyze added toxicity when translating a large evaluation dataset (HOLISTICBIAS, over 472k sentences, covering 13 demographic axes) from English into 164 languages. An automatic toxicity evaluation shows that added toxicity across languages varies from 0% to 5%. The output languages with the most added toxicity tend to be low-resource ones, and the demographic axes with the most added toxicity include sexual orientation, gender and sex, and ability. We also perform human evaluation on a subset of 8 translation directions, confirming the prevalence of true added toxicity. We use a measurement of the amount of source contribution to the translation, where a low source contribution implies hallucination, to interpret what causes toxicity. Making use of the input attributions allows us to explain toxicity, because the source contributions significantly correlate with toxicity for 84% of languages studied. Given our findings, our recommendations to reduce added toxicity are to curate training data to avoid mistranslations, mitigate hallucination and check unstable translations.

Published
2022

110. Supercooling of the A phase of $^3$He

Author

Tian, Yefan, Lotnyk, Dmytro, Eyal, Anna, Zhang, Kuang, Zhelev, Nikolay, Abhilash, T. S., Chavez, Aldo, Smith, Eric, Hindmarsh, Mark, Saunders, John, Mueller, Erich, and Parpia, Jeevak

Subjects
Condensed Matter - Other Condensed Matter, Condensed Matter - Mesoscale and Nanoscale Physics, Condensed Matter - Quantum Gases, Condensed Matter - Superconductivity
Abstract

Because of the extreme purity, lack of disorder, and complex order parameter, the first-order superfluid $^3$He A-B transition is the leading model system for first order transitions in the early universe. Here we report on the path dependence of the supercooling of the A phase over a wide range of pressures below 29.3 bar at nearly zero magnetic field. The A phase can be cooled significantly below the thermodynamic A-B transition temperature. While the extent of supercooling is highly reproducible, it depends strongly upon the cooling trajectory: The metastability of the A phase is enhanced by transiting through regions where the A phase is more stable. We provide evidence that some of the additional supercooling is due to the elimination of B phase seeds formed upon passage through the superfluid transition. A greater understanding of the physics is essential before the $^3$He can be exploited to model transitions in the early universe., Comment: 6 Figures

Published
2022
Full Text
View/download PDF

111. BlenderBot 3: a deployed conversational agent that continually learns to responsibly engage

Author

Shuster, Kurt, Xu, Jing, Komeili, Mojtaba, Ju, Da, Smith, Eric Michael, Roller, Stephen, Ung, Megan, Chen, Moya, Arora, Kushal, Lane, Joshua, Behrooz, Morteza, Ngan, William, Poff, Spencer, Goyal, Naman, Szlam, Arthur, Boureau, Y-Lan, Kambadur, Melanie, and Weston, Jason

Subjects
Computer Science - Computation and Language, Computer Science - Artificial Intelligence
Abstract

We present BlenderBot 3, a 175B parameter dialogue model capable of open-domain conversation with access to the internet and a long-term memory, and having been trained on a large number of user defined tasks. We release both the model weights and code, and have also deployed the model on a public web page to interact with organic users. This technical report describes how the model was built (architecture, model and training scheme), and details of its deployment, including safety mechanisms. Human evaluations show its superiority to existing open-domain dialogue agents, including its predecessors (Roller et al., 2021; Komeili et al., 2022). Finally, we detail our plan for continual learning using the data collected from deployment, which will also be publicly released. The goal of this research program is thus to enable the community to study ever-improving responsible agents that learn through interaction.

Published
2022

112. The Science Performance of JWST as Characterized in Commissioning

Author

Rigby, Jane, Perrin, Marshall, McElwain, Michael, Kimble, Randy, Friedman, Scott, Lallo, Matt, Doyon, René, Feinberg, Lee, Ferruit, Pierre, Glasse, Alistair, Rieke, Marcia, Rieke, George, Wright, Gillian, Willott, Chris, Colon, Knicole, Milam, Stefanie, Neff, Susan, Stark, Christopher, Valenti, Jeff, Abell, Jim, Abney, Faith, Abul-Huda, Yasin, Acton, D. Scott, Adams, Evan, Adler, David, Aguilar, Jonathan, Ahmed, Nasif, Albert, Loïc, Alberts, Stacey, Aldridge, David, Allen, Marsha, Altenburg, Martin, Marquez, Javier Alvarez, de Oliveira, Catarina Alves, Andersen, Greg, Anderson, Harry, Anderson, Sara, Argyriou, Ioannis, Armstrong, Amber, Arribas, Santiago, Artigau, Etienne, Arvai, Amanda, Atkinson, Charles, Bacon, Gregory, Bair, Thomas, Banks, Kimberly, Barrientes, Jaclyn, Barringer, Bruce, Bartosik, Peter, Bast, William, Baudoz, Pierre, Beatty, Thomas, Bechtold, Katie, Beck, Tracy, Bergeron, Eddie, Bergkoetter, Matthew, Bhatawdekar, Rachana, Birkmann, Stephan, Blazek, Ronald, Blome, Claire, Boccaletti, Anthony, Boeker, Torsten, Boia, John, Bonaventura, Nina, Bond, Nicholas, Bosley, Kari, Boucarut, Ray, Bourque, Matthew, Bouwman, Jeroen, Bower, Gary, Bowers, Charles, Boyer, Martha, Bradley, Larry, Brady, Greg, Braun, Hannah, Breda, David, Bresnahan, Pamela, Bright, Stacey, Britt, Christopher, Bromenschenkel, Asa, Brooks, Brian, Brooks, Keira, Brown, Bob, Brown, Matthew, Brown, Patricia, Bunker, Andy, Burger, Matthew, Bushouse, Howard, Cale, Steven, Cameron, Alex, Cameron, Peter, Canipe, Alicia, Caplinger, James, Caputo, Francis, Cara, Mihai, Carey, Larkin, Carniani, Stefano, Carrasquilla, Maria, Carruthers, Margaret, Case, Michael, Catherine, Riggs, Chance, Don, Chapman, George, Charlot, Stéphane, Charlow, Brian, Chayer, Pierre, Chen, Bin, Cherinka, Brian, Chichester, Sarah, Chilton, Zack, Chonis, Taylor, Clampin, Mark, Clark, Charles, Clark, Kerry, Coe, Dan, Coleman, Benee, Comber, Brian, Comeau, Tom, Connolly, Dennis, Cooper, James, Cooper, Rachel, Coppock, Eric, Correnti, Matteo, Cossou, Christophe, Coulais, Alain, Coyle, Laura, Cracraft, Misty, Curti, Mirko, Cuturic, Steven, Davis, Katherine, Davis, Michael, Dean, Bruce, DeLisa, Amy, deMeester, Wim, Dencheva, Nadia, Dencheva, Nadezhda, DePasquale, Joseph, Deschenes, Jeremy, Detre, Örs Hunor, Diaz, Rosa, Dicken, Dan, DiFelice, Audrey, Dillman, Matthew, Dixon, William, Doggett, Jesse, Donaldson, Tom, Douglas, Rob, DuPrie, Kimberly, Dupuis, Jean, Durning, John, Easmin, Nilufar, Eck, Weston, Edeani, Chinwe, Egami, Eiichi, Ehrenwinkler, Ralf, Eisenhamer, Jonathan, Eisenhower, Michael, Elie, Michelle, Elliott, James, Elliott, Kyle, Ellis, Tracy, Engesser, Michael, Espinoza, Nestor, Etienne, Odessa, Etxaluze, Mireya, Falini, Patrick, Feeney, Matthew, Ferry, Malcolm, Filippazzo, Joseph, Fincham, Brian, Fix, Mees, Flagey, Nicolas, Florian, Michael, Flynn, Jim, Fontanella, Erin, Ford, Terrance, Forshay, Peter, Fox, Ori, Franz, David, Fu, Henry, Fullerton, Alexander, Galkin, Sergey, Galyer, Anthony, Marin, Macarena Garcia, Gardner, Jonathan, Gardner, Lisa, Garland, Dennis, Garrett, Bruce, Gasman, Danny, Gaspar, Andras, Gaudreau, Daniel, Gauthier, Peter, Geers, Vincent, Geithner, Paul, Gennaro, Mario, Giardino, Giovanna, Girard, Julien, Giuliano, Mark, Glassmire, Kirk, Glauser, Adrian, Glazer, Stuart, Godfrey, John, Golimowski, David, Gollnitz, David, Gong, Fan, Gonzaga, Shireen, Gordon, Michael, Gordon, Karl, Goudfrooij, Paul, Greene, Thomas, Greenhouse, Matthew, Grimaldi, Stefano, Groebner, Andrew, Grundy, Timothy, Guillard, Pierre, Gutman, Irvin, Ha, Kong Q., Haderlein, Peter, Hagedorn, Andria, Hainline, Kevin, Haley, Craig, Hami, Maryam, Hamilton, Forrest, Hammel, Heidi, Hansen, Carl, Harkins, Tom, Harr, Michael, Hart, Jessica, Hart, Quyen, Hartig, George, Hashimoto, Ryan, Haskins, Sujee, Hathaway, William, Havey, Keith, Hayden, Brian, Hecht, Karen, Heller-Boyer, Chris, Henriques, Caroline, Henry, Alaina, Hermann, Karl, Hernandez, Scarlin, Hesman, Brigette, Hicks, Brian, Hilbert, Bryan, Hines, Dean, Hoffman, Melissa, Holfeltz, Sherie, Holler, Bryan J., Hoppa, Jennifer, Hott, Kyle, Howard, Joseph, Howard, Rick, Hunter, Alexander, Hunter, David, Hurst, Brendan, Husemann, Bernd, Hustak, Leah, Ignat, Luminita Ilinca, Illingworth, Garth, Irish, Sandra, Jackson, Wallace, Jahromi, Amir, Jakobsen, Peter, James, LeAndrea, James, Bryan, Januszewski, William, Jenkins, Ann, Jirdeh, Hussein, Johnson, Phillip, Johnson, Timothy, Jones, Vicki, Jones, Ron, Jones, Danny, Jones, Olivia, Jordan, Ian, Jordan, Margaret, Jurczyk, Sarah, Jurling, Alden, Kaleida, Catherine, Kalmanson, Phillip, Kammerer, Jens, Kang, Huijo, Kao, Shaw-Hong, Karakla, Diane, Kavanagh, Patrick, Kelly, Doug, Kendrew, Sarah, Kennedy, Herbert, Kenny, Deborah, Keski-kuha, Ritva, Keyes, Charles, Kidwell, Richard, Kinzel, Wayne, Kirk, Jeff, Kirkpatrick, Mark, Kirshenblat, Danielle, Klaassen, Pamela, Knapp, Bryan, Knight, J. Scott, Knollenberg, Perry, Koehler, Robert, Koekemoer, Anton, Kovacs, Aiden, Kulp, Trey, Kumari, Nimisha, Kyprianou, Mark, La Massa, Stephanie, Labador, Aurora, Ortega, Alvaro Labiano, Lagage, Pierre-Olivier, Lajoie, Charles-Phillipe, Lallo, Matthew, Lam, May, Lamb, Tracy, Lambros, Scott, Lampenfield, Richard, Langston, James, Larson, Kirsten, Law, David, Lawrence, Jon, Lee, David, Leisenring, Jarron, Lepo, Kelly, Leveille, Michael, Levenson, Nancy, Levine, Marie, Levy, Zena, Lewis, Dan, Lewis, Hannah, Libralato, Mattia, Lightsey, Paul, Link, Miranda, Liu, Lily, Lo, Amy, Lockwood, Alexandra, Logue, Ryan, Long, Chris, Long, Douglas, Loomis, Charles, Lopez-Caniego, Marcos, Alvarez, Jose Lorenzo, Love-Pruitt, Jennifer, Lucy, Adrian, Luetzgendorf, Nora, Maghami, Peiman, Maiolino, Roberto, Major, Melissa, Malla, Sunita, Malumuth, Eliot, Manjavacas, Elena, Mannfolk, Crystal, Marrione, Amanda, Marston, Anthony, Martel, André, Maschmann, Marc, Masci, Gregory, Masciarelli, Michaela, Maszkiewicz, Michael, Mather, John, McKenzie, Kenny, McLean, Brian, McMaster, Matthew, Melbourne, Katie, Meléndez, Marcio, Menzel, Michael, Merz, Kaiya, Meyett, Michele, Meza, Luis, Miskey, Cherie, Misselt, Karl, Moller, Christopher, Morrison, Jane, Morse, Ernie, Moseley, Harvey, Mosier, Gary, Mountain, Matt, Mueckay, Julio, Mueller, Michael, Mullally, Susan, Murphy, Jess, Murray, Katherine, Murray, Claire, Mustelier, David, Muzerolle, James, Mycroft, Matthew, Myers, Richard, Myrick, Kaila, Nanavati, Shashvat, Nance, Elizabeth, Nayak, Omnarayani, Naylor, Bret, Nelan, Edmund, Nickson, Bryony, Nielson, Alethea, Nieto-Santisteban, Maria, Nikolov, Nikolay, Noriega-Crespo, Alberto, O'Shaughnessy, Brian, O'Sullivan, Brian, Ochs, William, Ogle, Patrick, Oleszczuk, Brenda, Olmsted, Joseph, Osborne, Shannon, Ottens, Richard, Owens, Beverly, Pacifici, Camilla, Pagan, Alyssa, Page, James, Park, Sang, Parrish, Keith, Patapis, Polychronis, Paul, Lee, Pauly, Tyler, Pavlovsky, Cheryl, Pedder, Andrew, Peek, Matthew, Pena-Guerrero, Maria, Pennanen, Konstantin, Perez, Yesenia, Perna, Michele, Perriello, Beth, Phillips, Kevin, Pietraszkiewicz, Martin, Pinaud, Jean-Paul, Pirzkal, Norbert, Pitman, Joseph, Piwowar, Aidan, Platais, Vera, Player, Danielle, Plesha, Rachel, Pollizi, Joe, Polster, Ethan, Pontoppidan, Klaus, Porterfield, Blair, Proffitt, Charles, Pueyo, Laurent, Pulliam, Christine, Quirt, Brian, Neira, Irma Quispe, Alarcon, Rafael Ramos, Ramsay, Leah, Rapp, Greg, Rapp, Robert, Rauscher, Bernard, Ravindranath, Swara, Rawle, Timothy, Regan, Michael, Reichard, Timothy A., Reis, Carl, Ressler, Michael E., Rest, Armin, Reynolds, Paul, Rhue, Timothy, Richon, Karen, Rickman, Emily, Ridgaway, Michael, Ritchie, Christine, Rix, Hans-Walter, Robberto, Massimo, Robinson, Gregory, Robinson, Michael, Robinson, Orion, Rock, Frank, Rodriguez, David, Del Pino, Bruno Rodriguez, Roellig, Thomas, Rohrbach, Scott, Roman, Anthony, Romelfanger, Fred, Rose, Perry, Roteliuk, Anthony, Roth, Marc, Rothwell, Braden, Rowlands, Neil, Roy, Arpita, Royer, Pierre, Royle, Patricia, Rui, Chunlei, Rumler, Peter, Runnels, Joel, Russ, Melissa, Rustamkulov, Zafar, Ryden, Grant, Ryer, Holly, Sabata, Modhumita, Sabatke, Derek, Sabbi, Elena, Samuelson, Bridget, Sapp, Benjamin, Sappington, Bradley, Sargent, B., Sauer, Arne, Scheithauer, Silvia, Schlawin, Everett, Schlitz, Joseph, Schmitz, Tyler, Schneider, Analyn, Schreiber, Jürgen, Schulze, Vonessa, Schwab, Ryan, Scott, John, Sembach, Kenneth, Shanahan, Clare, Shaughnessy, Bryan, Shaw, Richard, Shawger, Nanci, Shay, Christopher, Sheehan, Evan, Shen, Sharon, Sherman, Allan, Shiao, Bernard, Shih, Hsin-Yi, Shivaei, Irene, Sienkiewicz, Matthew, Sing, David, Sirianni, Marco, Sivaramakrishnan, Anand, Skipper, Joy, Sloan, Gregory, Slocum, Christine, Slowinski, Steven, Smith, Erin, Smith, Eric, Smith, Denise, Smith, Corbett, Snyder, Gregory, Soh, Warren, Sohn, Tony, Soto, Christian, Spencer, Richard, Stallcup, Scott, Stansberry, John, Starr, Carl, Starr, Elysia, Stewart, Alphonso, Stiavelli, Massimo, Straughn, Amber, Strickland, David, Stys, Jeff, Summers, Francis, Sun, Fengwu, Sunnquist, Ben, Swade, Daryl, Swam, Michael, Swaters, Robert, Swoish, Robby, Taylor, Joanna M., Taylor, Rolanda, Plate, Maurice Te, Tea, Mason, Teague, Kelly, Telfer, Randal, Temim, Tea, Thatte, Deepashri, Thompson, Christopher, Thompson, Linda, Thomson, Shaun, Tikkanen, Tuomo, Tippet, William, Todd, Connor, Toolan, Sharon, Tran, Hien, Trejo, Edwin, Truong, Justin, Tsukamoto, Chris, Tustain, Samuel, Tyra, Harrison, Ubeda, Leonardo, Underwood, Kelli, Uzzo, Michael, Van Campen, Julie, Vandal, Thomas, Vandenbussche, Bart, Vila, Begoña, Volk, Kevin, Wahlgren, Glenn, Waldman, Mark, Walker, Chanda, Wander, Michel, Warfield, Christine, Warner, Gerald, Wasiak, Matthew, Watkins, Mitchell, Weaver, Andrew, Weilert, Mark, Weiser, Nick, Weiss, Ben, Weissman, Sarah, Welty, Alan, West, Garrett, Wheate, Lauren, Wheatley, Elizabeth, Wheeler, Thomas, White, Rick, Whiteaker, Kevin, Whitehouse, Paul, Whiteleather, Jennifer, Whitman, William, Williams, Christina, Willmer, Christopher, Willoughby, Scott, Wilson, Andrew, Wirth, Gregory, Wislowski, Emily, Wolf, Erin, Wolfe, David, Wolff, Schuyler, Workman, Bill, Wright, Ray, Wu, Carl, Wu, Rai, Wymer, Kristen, Yates, Kayla, Yeager, Christopher, Yeates, Jared, Yerger, Ethan, Yoon, Jinmi, Young, Alice, Yu, Susan, Zak, Dean, Zeidler, Peter, Zhou, Julia, Zielinski, Thomas, Zincke, Cristian, and Zonak, Stephanie

Subjects
Astrophysics - Instrumentation and Methods for Astrophysics
Abstract

This paper characterizes the actual science performance of the James Webb Space Telescope (JWST), as determined from the six month commissioning period. We summarize the performance of the spacecraft, telescope, science instruments, and ground system, with an emphasis on differences from pre-launch expectations. Commissioning has made clear that JWST is fully capable of achieving the discoveries for which it was built. Moreover, almost across the board, the science performance of JWST is better than expected; in most cases, JWST will go deeper faster than expected. The telescope and instrument suite have demonstrated the sensitivity, stability, image quality, and spectral range that are necessary to transform our understanding of the cosmos through observations spanning from near-earth asteroids to the most distant galaxies., Comment: 5th version as accepted to PASP; 31 pages, 18 figures; https://iopscience.iop.org/article/10.1088/1538-3873/acb293

Published
2022
Full Text
View/download PDF

113. Spatial distributions of white matter hyperintensities on brain MRI: A pooled analysis of individual participant data from 11 memory clinic cohorts

Author

Coenen, Mirthe, Biessels, Geert Jan, DeCarli, Charles, Fletcher, Evan F, Maillard, Pauline M, Initiative, Alzheimer's Disease Neuroimaging, Barkhof, Frederik, Barnes, Josephine, Benke, Thomas, Boomsma, Jooske MF, Chen, Christopher PLH, Dal-Bianco, Peter, Dewenter, Anna, Duering, Marco, Enzinger, Christian, Ewers, Michael, Exalto, Lieza G, Franzmeier, Nicolai, Groeneveld, Onno, Hilal, Saima, Hofer, Edith, Koek, Huiberdina L, Maier, Andrea B, McCreary, Cheryl R, Papma, Janne M, Paterson, Ross W, Pijnenburg, Yolande AL, Rubinski, Anna, Schmidt, Reinhold, Schott, Jonathan M, Slattery, Catherine F, Smith, Eric E, Sudre, Carole H, Steketee, Rebecca ME, van den Berg, Esther, van der Flier, Wiesje M, Venketasubramanian, Narayanaswamy, Vernooij, Meike W, Wolters, Frank J, Xin, Xu, Biesbroek, J Matthijs, and Kuijf, Hugo J

Subjects
Biomedical and Clinical Sciences, Clinical Sciences, Alzheimer's Disease Related Dementias (ADRD), Cerebrovascular, Alzheimer's Disease including Alzheimer's Disease Related Dementias (AD/ADRD), Machine Learning and Artificial Intelligence, Clinical Research, Acquired Cognitive Impairment, Vascular Cognitive Impairment/Dementia, Aging, Neurodegenerative, Networking and Information Technology R&D (NITRD), Neurosciences, Brain Disorders, Dementia, Neurological, Humans, White Matter, Brain, Magnetic Resonance Imaging, Neuroimaging, Cognitive Dysfunction, Multicenter Studies as Topic, White matter hyperintensities, Brain MRI, Distribution frequencies, Lesion location, Alzheimer's Disease Neuroimaging Initiative, Biological psychology, Clinical and health psychology
Abstract

IntroductionThe spatial distribution of white matter hyperintensities (WMH) on MRI is often considered in the diagnostic evaluation of patients with cognitive problems. In some patients, clinicians may classify WMH patterns as "unusual", but this is largely based on expert opinion, because detailed quantitative information about WMH distribution frequencies in a memory clinic setting is lacking. Here we report voxel wise 3D WMH distribution frequencies in a large multicenter dataset and also aimed to identify individuals with unusual WMH patterns.MethodsIndividual participant data (N = 3525, including 777 participants with subjective cognitive decline, 1389 participants with mild cognitive impairment and 1359 patients with dementia) from eleven memory clinic cohorts, recruited through the Meta VCI Map Consortium, were used. WMH segmentations were provided by participating centers or performed in Utrecht and registered to the Montreal Neurological Institute (MNI)-152 brain template for spatial normalization. To determine WMH distribution frequencies, we calculated WMH probability maps at voxel level. To identify individuals with unusual WMH patterns, region-of-interest (ROI) based WMH probability maps, rule-based scores, and a machine learning method (Local Outlier Factor (LOF)), were implemented.ResultsWMH occurred in 82% of voxels from the white matter template with large variation between subjects. Only a small proportion of the white matter (1.7%), mainly in the periventricular areas, was affected by WMH in at least 20% of participants. A large portion of the total white matter was affected infrequently. Nevertheless, 93.8% of individual participants had lesions in voxels that were affected in less than 2% of the population, mainly located in subcortical areas. Only the machine learning method effectively identified individuals with unusual patterns, in particular subjects with asymmetric WMH distribution or with WMH at relatively rarely affected locations despite common locations not being affected.DiscussionAggregating data from several memory clinic cohorts, we provide a detailed 3D map of WMH lesion distribution frequencies, that informs on common as well as rare localizations. The use of data-driven analysis with LOF can be used to identify unusual patterns, which might serve as an alert that rare causes of WMH should be considered.

Published
2023

124. Perturbation Augmentation for Fairer NLP

Author

Qian, Rebecca, Ross, Candace, Fernandes, Jude, Smith, Eric, Kiela, Douwe, and Williams, Adina

Subjects
Computer Science - Computation and Language, Computer Science - Artificial Intelligence
Abstract

Unwanted and often harmful social biases are becoming ever more salient in NLP research, affecting both models and datasets. In this work, we ask whether training on demographically perturbed data leads to fairer language models. We collect a large dataset of human annotated text perturbations and train a neural perturbation model, which we show outperforms heuristic alternatives. We find that (i) language models (LMs) pre-trained on demographically perturbed corpora are typically more fair, and (ii) LMs finetuned on perturbed GLUE datasets exhibit less demographic bias on downstream tasks, and (iii) fairness improvements do not come at the expense of performance on downstream tasks. Lastly, we discuss outstanding questions about how best to evaluate the (un)fairness of large language models. We hope that this exploration of neural demographic perturbation will help drive more improvement towards fairer NLP.

Published
2022

125. 'I'm sorry to hear that': Finding New Biases in Language Models with a Holistic Descriptor Dataset

Author

Smith, Eric Michael, Hall, Melissa, Kambadur, Melanie, Presani, Eleonora, and Williams, Adina

Subjects
Computer Science - Computation and Language, Computer Science - Computers and Society
Abstract

As language models grow in popularity, it becomes increasingly important to clearly measure all possible markers of demographic identity in order to avoid perpetuating existing societal harms. Many datasets for measuring bias currently exist, but they are restricted in their coverage of demographic axes and are commonly used with preset bias tests that presuppose which types of biases models can exhibit. In this work, we present a new, more inclusive bias measurement dataset, HolisticBias, which includes nearly 600 descriptor terms across 13 different demographic axes. HolisticBias was assembled in a participatory process including experts and community members with lived experience of these terms. These descriptors combine with a set of bias measurement templates to produce over 450,000 unique sentence prompts, which we use to explore, identify, and reduce novel forms of bias in several generative models. We demonstrate that HolisticBias is effective at measuring previously undetectable biases in token likelihoods from language models, as well as in an offensiveness classifier. We will invite additions and amendments to the dataset, which we hope will serve as a basis for more easy-to-use and standardized methods for evaluating bias in NLP models., Comment: EMNLP 2022

Published
2022

126. US Surveillance of Acute Ischemic Stroke Patient Characteristics, Care Quality, and Outcomes for 2019.

Author

Ziaeian, Boback, Xu, Haolin, Matsouaka, Roland A, Xian, Ying, Khan, Yosef, Schwamm, Lee S, Smith, Eric E, and Fonarow, Gregg C

Subjects
Humans, Fibrinolytic Agents, Anticoagulants, Platelet Aggregation Inhibitors, Hydroxymethylglutaryl-CoA Reductase Inhibitors, Treatment Outcome, Bayes Theorem, Aged, Quality of Health Care, United States, Female, Male, Stroke, Ischemic Stroke, Bayesian analysis, epidemiology, health services, ischemic stroke, quality and outcomes, Prevention, Health Services, Clinical Research, Brain Disorders, Good Health and Well Being, Cardiorespiratory Medicine and Haematology, Clinical Sciences, Neurosciences, Neurology & Neurosurgery
Abstract

BackgroundThe United States lacks a timely and accurate nationwide surveillance system for acute ischemic stroke (AIS). We use the Get With The Guidelines-Stroke registry to apply poststratification survey weights to generate national assessment of AIS epidemiology, hospital care quality, and in-hospital outcomes.MethodsClinical data from the Get With The Guidelines-Stroke registry were weighted using a Bayesian interpolation method anchored to observations from the national inpatient sample. To generate a US stroke forecast for 2019, we linearized time trend estimates from the national inpatient sample to project anticipated AIS hospital volume, distribution, and race/ethnicity characteristics for the year 2019. Primary measures of AIS epidemiology and clinical care included patient and hospital characteristics, stroke severity, vital and laboratory measures, treatment interventions, performance measures, disposition, and clinical outcomes at discharge.ResultsWe estimate 552 476 patients with AIS were admitted in 2019 to US hospitals. Median age was 71 (interquartile range, 60-81), 48.8% female. Atrial fibrillation was diagnosed in 22.6%, 30.2% had prior stroke/transient ischemic attack, and 36.4% had diabetes. At baseline, 46.4% of patients with AIS were taking antiplatelet agents, 19.2% anticoagulants, and 46.3% cholesterol-reducers. Mortality was 4.4%, and only 52.3% were able to ambulate independently at discharge. Performance nationally on AIS achievement measures were generally higher than 95% for all measures but the use of thrombolytics within 3 hours of early stroke presentations (81.9%). Additional quality measures had lower rates of receipt: dysphagia screening (84.9%), early thrombolytics by 4.5 hours (79.7%), and statin therapy (80.6%).ConclusionsWe provide timely, reliable, and actionable US national AIS surveillance using Bayesian interpolation poststratification weights. These data may facilitate more targeted quality improvement efforts, resource allocation, and national policies to improve AIS care and outcomes.

Published
2022

127. Outcomes of Arthroscopic Lysis of Adhesions for the Treatment of Postoperative Knee Arthrofibrosis: A Systematic Review

Author

Fackler, Nathan, Chin, Garwin, Karasavvidis, Theofilos, Bohlen, Hunter, Smith, Eric, Amirhekmat, Arya, and Wang, Dean

Subjects
Biomedical and Clinical Sciences, Clinical Sciences, Patient Safety, Evaluation of treatments and therapeutic interventions, 6.4 Surgery, 6.1 Pharmaceuticals, arthrofibrosis, arthroscopy, lysis of adhesions, postoperative, Human Movement and Sports Sciences, Clinical sciences, Sports science and exercise
Abstract

BackgroundPostoperative knee arthrofibrosis is a common and potentially detrimental complication affecting knee function and gait. Several cohort studies have reported good outcomes after arthroscopic lysis of adhesions (LOA) with manipulation under anesthesia (MUA).PurposeTo review the literature assessing the efficacy and complications of arthroscopic LOA and MUA for postoperative arthrofibrosis of the knee and evaluate whether any relevant subgroups are associated with different clinical presentation and outcomes.Study designSystematic review; Level of evidence, 4.MethodsThis review was performed according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. Eligible studies published from January 1, 1990, to April 1, 2021, were identified through a search of the US National Library of Medicine (PubMed/MEDLINE), EMBASE, and Cochrane databases. All studies included in this analysis included pre- and postoperative range of motion measurements for their treated patients. Studies reporting outcomes for patients with isolated cyclops lesions after anterior cruciate ligament reconstruction were excluded.ResultsEight studies comprising 240 patients were included. The mean time from index surgery to arthroscopic LOA and MUA was 8.4 months, and the mean postoperative follow-up was at 31.2 months. All studies demonstrated a significant improvement (41.6°) in arc of motion after arthroscopic LOA. Clinically significant improvements in outcome measures, including the International Knee Documentation Committee, Western Ontario and McMaster Universities Osteoarthritis Index, and Knee injury and Osteoarthritis Outcome Score, were reported after arthroscopic LOA across all applicable studies. Of 240 patients, a single complication (synovial fistula) occurred after LOA and MUA, which resolved without intervention.ConclusionThe results of this review indicated that arthroscopic LOA and MUA is a safe and efficacious treatment for postoperative arthrofibrosis of the knee.

Published
2022

128. Brain herniation (encephalocele) into arachnoid granulations: prevalence and association with pulsatile tinnitus and idiopathic intracranial hypertension.

Author

Smith, Eric R, Caton, M Travis, Villanueva-Meyer, Javier E, Remer, Justin, Eisenmenger, Laura B, Baker, Amanda, Shah, Vinil N, Tu-Chan, Adelyn, Meisel, Karl, and Amans, Matthew R

Subjects
Cranial Sinuses, Brain, Arachnoid, Humans, Tinnitus, Brain Diseases, Intracranial Hypertension, Pseudotumor Cerebri, Encephalocele, Prevalence, Retrospective Studies, Arachnoid granulation, Brain herniation, Idiopathic intracranial hypertension, Magnetic resonance imaging, Pulsatile tinnitus, Biomedical Imaging, Neurosciences, Brain Disorders, Prevention, Clinical Sciences, Nuclear Medicine & Medical Imaging
Abstract

PurposeBrain herniation into arachnoid granulations (BHAG) of the dural venous sinuses is a recently described finding of uncertain etiology. The purpose of this study was to investigate the prevalence of BHAG in a cohort of patients with pulsatile tinnitus (PT) and to clarify the physiologic and clinical implications of these lesions.MethodsThe imaging and charts of consecutive PT patients were retrospectively reviewed. All patients were examined with MRI including pre- and post-contrast T1- and T2-weighted sequences. Images were reviewed separately by three blinded neuroradiologists to identify the presence of BHAG. Their location, signal intensity, size, presence of arachnoid granulation, and associated dural venous sinus stenosis were documented. Clinical records were further reviewed for idiopathic intracranial hypertension, history of prior lumbar puncture, and opening pressure.ResultsTwo hundred sixty-two consecutive PT patients over a 4-year period met inclusion criteria. PT patients with BHAG were significantly more likely to have idiopathic intracranial hypertension than PT patients without BHAG (OR 4.2, CI 1.5-12, p = 0.006). Sixteen out of 262 (6%) patients were found to have 18 BHAG. Eleven out of 16 (69%) patients had unilateral temporal or occipital lobe herniations located in the transverse sinus or the transverse-sigmoid junction. Three out of 16 (19%) patients had unilateral cerebellar herniations and 2/16 (13%) patients had bilateral BHAG.ConclusionIn patients with PT, BHAG is a prevalent MRI finding that is strongly associated with the clinical diagnosis of IIH. The pathogenesis of BHAG remains uncertain, but recognition should prompt comprehensive evaluation for IIH.

Published
2022

131. A PD-1-targeted, receptor-masked IL-2 immunocytokine that engages IL-2Rα strengthens T cell-mediated anti-tumor therapies

Author

Wu, Jiaxi, Bloch, Nicolin, Chang, Aaron Y., Bhavsar, Ramandeep, Wang, Qingqing, Crawford, Alison, DiLillo, David J., Vazzana, Kristin, Mohrs, Katja, Dudgeon, Drew, Patel, Supriya, Ahmed, Hassan, Garg, Vidur, Amatulli, Michael, Antao, Olivia Q., Yan, Yuetian, Wang, Shunhai, Ramos, Willy, Krueger, Pamela, Adler, Christina, Ni, Min, Wei, Yi, Guo, Chunguang, Macdonald, Lynn, Huang, Tammy, Ullman, Erica, Hermann, Aynur, Yancopoulos, George D., Murphy, Andrew J., Davis, Samuel, Olson, William C., Lin, John C., Smith, Eric, and Zhang, Tong

Published
2024
Full Text
View/download PDF

135. Human Evaluation of Conversations is an Open Problem: comparing the sensitivity of various methods for evaluating dialogue agents

Author

Smith, Eric Michael, Hsu, Orion, Qian, Rebecca, Roller, Stephen, Boureau, Y-Lan, and Weston, Jason

Subjects
Computer Science - Computation and Language, Computer Science - Artificial Intelligence
Abstract

At the heart of improving conversational AI is the open problem of how to evaluate conversations. Issues with automatic metrics are well known (Liu et al., 2016, arXiv:1603.08023), with human evaluations still considered the gold standard. Unfortunately, how to perform human evaluations is also an open problem: differing data collection methods have varying levels of human agreement and statistical sensitivity, resulting in differing amounts of human annotation hours and labor costs. In this work we compare five different crowdworker-based human evaluation methods and find that different methods are best depending on the types of models compared, with no clear winner across the board. While this highlights the open problems in the area, our analysis leads to advice of when to use which one, and possible future directions.

Published
2022

140. Proceedings from the Blood and Marrow Transplant Clinical Trials Network Myeloma Intergroup Workshop on Immune and Cellular Therapy in Multiple Myeloma.

Author

Holstein, Sarah, Asimakopoulos, Fotis, Azab, Abdel, Bianchi, Giada, Bhutani, Manisha, Crews, Leslie, Cupedo, Tom, Giles, Hannah, Gooding, Sarah, Hillengass, Jens, John, Lukas, Kaiser, Shari, Lee, Lydia, Maclachlan, Kylee, Pasquini, Marcelo, Pichiorri, Flavia, Shokeen, Monica, Shy, Brian, Smith, Eric, Verona, Raluca, Usmani, Saad, McCarthy, Philip, and Shah, Nina

Subjects
CAR T cell, Imaging, Multiple myeloma, Resistance, Tumor microenvironment, Bone Marrow, Cell- and Tissue-Based Therapy, Clinical Trials as Topic, Humans, Multiple Myeloma
Abstract

The Blood and Marrow Transplant Clinical Trials Network (BMT CTN) Myeloma Intergroup conducted a workshop on Immune and Cellular Therapy in Multiple Myeloma on January 7, 2022. This workshop included presentations by basic, translational, and clinical researchers with expertise in plasma cell dyscrasias. Four main topics were discussed: platforms for myeloma disease evaluation, insights into pathophysiology, therapeutic target and resistance mechanisms, and cellular therapy for multiple myeloma. Here we provide a comprehensive summary of these workshop presentations.

Published
2022

141. Transradial Approach for Thoracolumbar Spinal Angiography and Tumor Embolization: Feasibility and Technical Considerations

Author

Caton, Michael Travis, Smith, Eric Robert, Baker, Amanda, Dowd, Christopher Foley, and Higashida, Randall T

Subjects
Biomedical and Clinical Sciences, Clinical Sciences, Endovascular procedures, Spinal cord vascular diseases, Radial artery, Angiography, digital subtraction, Transradial approach, Angiography, digital subtraction, Clinical sciences, Neurosciences
Abstract

The transradial approach (TRA) is an effective and safe alternative to transfemoral access for diagnostic neuroangiography and craniocervical interventions. While the technical aspects of supraclavicular intervention are well-described, there are little data on the TRA for thoracolumbar angiography and intervention. The authors describe the feasibility of the TRA for preoperative thoracic tumor embolization, emphasizing technique, device selection, navigation, and catheterization of thoracolumbar segmental arteries. This approach extends the benefits of TRA to spinal interventional neuroradiology.

Published
2022

142. Hi, my name is Martha: Using names to measure and mitigate bias in generative dialogue models

Author

Smith, Eric Michael and Williams, Adina

Subjects
Computer Science - Computation and Language
Abstract

All AI models are susceptible to learning biases in data that they are trained on. For generative dialogue models, being trained on real human conversations containing unbalanced gender and race/ethnicity references can lead to models that display learned biases, which we define here broadly as any measurable differences in the distributions of words or semantic content of conversations based on demographic groups. We measure the strength of such biases by producing artificial conversations between two copies of a dialogue model, conditioning one conversational partner to state a name commonly associated with a certain gender and/or race/ethnicity. We find that larger capacity models tend to exhibit more gender bias and greater stereotyping of occupations by gender. We show that several methods of tuning these dialogue models, specifically name scrambling, controlled generation, and unlikelihood training, are effective in reducing bias in conversation, including on a downstream conversational task. Name scrambling is also effective in lowering differences in token usage across conversations where partners have names associated with different genders or races/ethnicities.

Published
2021

145. Activity limitations, use of assistive devices, and mortality and clinical events in 25 high-income, middle-income, and low-income countries: an analysis of the PURE study

Author

Joundi, Raed A, Hu, Bo, Rangarajan, Sumathy, Leong, Darryl P, Islam, Shofiqul, Smith, Eric E, Mirrakhimov, Erkin, Seron, Pamela, Alhabib, Khalid F, Assembekov, Batyrbek, Chifamba, Jephat, Yusuf, Rita, Khatib, Rasha, Felix, Camilo, Yusufali, Afzalhussein, Mohammadifard, Noushin, Rosengren, Annika, Oguz, Aytekin, Iqbal, Romaina, Yeates, Karen, Avezum, Alvaro, Kruger, Iolanthé, Anjana, Ranjit, PVM, Lakshmi, Gupta, Rajeev, Zatońska, Katarzyna, Barbarash, Olga, Pelliza, Eugenia, Rammohan, Kamala, Li, Mengya, Li, Xiaocong, Ismail, Rosnah, Lopez-Jaramillo, Patricio, Evans, Marc, O'Donnell, Martin, and Yusuf, Salim

Published
2024
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results