Search

Autonomous agents powered by large language models (LLMs) have attracted significant research interest. However, the open-source community faces many challenges in developing specialized models for agent tasks, driven by the scarcity of high-quality agent datasets and the absence of standard protocols in this area. We introduce and publicly release xLAM, a series of large action models designed for AI agent tasks. The xLAM series includes five models with both dense and mixture-of-expert architectures, ranging from 1B to 8x22B parameters, trained using a scalable, flexible pipeline that unifies, augments, and synthesizes diverse datasets to enhance AI agents' generalizability and performance across varied environments. Our experimental results demonstrate that xLAM consistently delivers exceptional performance across multiple agent ability benchmarks, notably securing the 1st position on the Berkeley Function-Calling Leaderboard, outperforming GPT-4, Claude-3, and many other models in terms of tool use. By releasing the xLAM series, we aim to advance the performance of open-source LLMs for autonomous AI agents, potentially accelerating progress and democratizing access to high-performance models for agent tasks. Models are available at https://huggingface.co/collections/Salesforce/xlam-models-65f00e2a0a63bbcd1c2dade4, Comment: Technical report for the Salesforce xLAM model series

The advancement of function-calling agent models requires diverse, reliable, and high-quality datasets. This paper presents APIGen, an automated data generation pipeline designed to synthesize verifiable high-quality datasets for function-calling applications. We leverage APIGen and collect 3,673 executable APIs across 21 different categories to generate diverse function-calling datasets in a scalable and structured manner. Each data in our dataset is verified through three hierarchical stages: format checking, actual function executions, and semantic verification, ensuring its reliability and correctness. We demonstrate that models trained with our curated datasets, even with only 7B parameters, can achieve state-of-the-art performance on the Berkeley Function-Calling Benchmark, outperforming multiple GPT-4 models. Moreover, our 1B model achieves exceptional performance, surpassing GPT-3.5-Turbo and Claude-3 Haiku. We release a dataset containing 60,000 high-quality entries, aiming to advance the field of function-calling agent domains. The dataset is available on Huggingface: https://huggingface.co/datasets/Salesforce/xlam-function-calling-60k and the project homepage: https://apigen-pipeline.github.io/

Our study investigates the impact of data augmentation on the performance of multivariate time series models, focusing on datasets from the UCR archive. Despite the limited size of these datasets, we achieved classification accuracy improvements in 10 out of 13 datasets using the Rocket and InceptionTime models. This highlights the essential role of sufficient data in training effective models, paralleling the advancements seen in computer vision. Our work delves into adapting and applying existing methods in innovative ways to the domain of multivariate time series classification. Our comprehensive exploration of these techniques sets a new standard for addressing data scarcity in time series analysis, emphasizing that diverse augmentation strategies are crucial for unlocking the potential of both traditional and deep learning models. Moreover, by meticulously analyzing and applying a variety of augmentation techniques, we demonstrate that strategic data enrichment can enhance model accuracy. This not only establishes a benchmark for future research in time series analysis but also underscores the importance of adopting varied augmentation approaches to improve model performance in the face of limited data availability., Comment: Workshop on Multivariate Time Series Analytics (MulTiSA), ICDE Workshop

This research paper presents an approach to enhancing the predictive capability of architects in the design and assurance of systems, focusing on systems operating in dynamic and unpredictable environments. By adopting a systems approach, we aim to improve architects' predictive capabilities in designing dependable systems (for example, ML-based systems). An aerospace case study is used to illustrate the approach. Multiple factors (challenges) influencing aircraft detection are identified, demonstrating the effectiveness of our approach in a complex operational setting. Our approach primarily aimed to enhance the architect's predictive capability., Comment: S. Bernardi, T. Zoppi (Editors), "Fast Abstracts and Student Forum Proceedings - EDCC 2024 - 19th European Dependable Computing Conference, Leuven, Belgium, 8-11 April 2024"

Autonomous agents powered by large language models (LLMs) have garnered significant research attention. However, fully harnessing the potential of LLMs for agent-based tasks presents inherent challenges due to the heterogeneous nature of diverse data sources featuring multi-turn trajectories. In this paper, we introduce \textbf{AgentOhana} as a comprehensive solution to address these challenges. \textit{AgentOhana} aggregates agent trajectories from distinct environments, spanning a wide array of scenarios. It meticulously standardizes and unifies these trajectories into a consistent format, streamlining the creation of a generic data loader optimized for agent training. Leveraging the data unification, our training pipeline maintains equilibrium across different data sources and preserves independent randomness across devices during dataset partitioning and model training. Additionally, we present \textbf{xLAM-v0.1}, a large action model tailored for AI agents, which demonstrates exceptional performance across various benchmarks. Begin the exploration at \url{https://github.com/SalesforceAIResearch/xLAM}., Comment: Add GitHub repo link at \url{https://github.com/SalesforceAIResearch/xLAM} and HuggingFace model link at \url{https://huggingface.co/Salesforce/xLAM-v0.1-r}

Balancing the trade-off between accuracy and robustness is a long-standing challenge in time series forecasting. While most of existing robust algorithms have achieved certain suboptimal performance on clean data, sustaining the same performance level in the presence of data perturbations remains extremely hard. In this paper, we study a wide array of perturbation scenarios and propose novel defense mechanisms against adversarial attacks using real-world telecom data. We compare our strategy against two existing adversarial training algorithms under a range of maximal allowed perturbations, defined using $\ell_{\infty}$-norm, $\in [0.1,0.4]$. Our findings reveal that our hybrid strategy, which is composed of a classifier to detect adversarial examples, a denoiser to eliminate noise from the perturbed data samples, and a standard forecaster, achieves the best performance on both clean and perturbed data. Our optimal model can retain up to $92.02\%$ the performance of the original forecasting model in terms of Mean Squared Error (MSE) on clean data, while being more robust than the standard adversarially trained models on perturbed data. Its MSE is 2.71$\times$ and 2.51$\times$ lower than those of comparing methods on normal and perturbed data, respectively. In addition, the components of our models can be trained in parallel, resulting in better computational efficiency. Our results indicate that we can optimally balance the trade-off between the performance and robustness of forecasting models by improving the classifier and denoiser, even in the presence of sophisticated and destructive poisoning attacks., Comment: Accepted for presentation at the ARTMAN workshop, part of the ACM Conference on Computer and Communications Security (CCS), 2023

How to identify those equivalent entities between knowledge graphs (KGs), which is called Entity Alignment (EA), is a long-standing challenge. So far, many methods have been proposed, with recent focus on leveraging Deep Learning to solve this problem. However, we observe that most of the efforts has been paid to having better representation of entities, rather than improving entity matching from the learned representations. In fact, how to efficiently infer the entity pairs from this similarity matrix, which is essentially a matching problem, has been largely ignored by the community. Motivated by this observation, we conduct an in-depth analysis on existing algorithms that are particularly designed for solving this matching problem, and propose a novel matching method, named Bidirectional Matching (BMat). Our extensive experimental results on public datasets indicate that there is currently no single silver bullet solution for EA. In other words, different classes of entity similarity estimation may require different matching algorithms to reach the best EA results for each class. We finally conclude that using PARIS, the state-of-the-art EA approach, with BMat gives the best combination in terms of EA performance and the algorithm's time and space complexity., Comment: 11 pages, 1 figure, 7 tables

Wildfire forecasting has been one of the most critical tasks that humanities want to thrive. It plays a vital role in protecting human life. Wildfire prediction, on the other hand, is difficult because of its stochastic and chaotic properties. We tackled the problem by interpreting a series of wildfire images as a video and used it to anticipate how the fire would behave in the future. However, creating video prediction models that account for the inherent uncertainty of the future is challenging. The bulk of published attempts is based on stochastic image-autoregressive recurrent networks, which raises various performance and application difficulties, such as computational cost and limited efficiency on massive datasets. Another possibility is to use entirely latent temporal models that combine frame synthesis and temporal dynamics. However, due to design and training issues, no such model for stochastic video prediction has yet been proposed in the literature. This paper addresses these issues by introducing a novel stochastic temporal model whose dynamics are driven in a latent space. It naturally predicts video dynamics by allowing our lighter, more interpretable latent model to beat previous state-of-the-art approaches on the GOES-16 dataset. Results will be compared towards various benchmarking models., Comment: arXiv admin note: substantial text overlap with arXiv:2002.09219 by other authors

Key Performance Indicators (KPI), which are essentially time series data, have been widely used to indicate the performance of telecom networks. Based on the given KPIs, a large set of anomaly detection algorithms have been deployed for detecting the unexpected network incidents. Generally, unsupervised anomaly detection algorithms gain more popularity than the supervised ones, due to the fact that labeling KPIs is extremely time- and resource-consuming, and error-prone. However, those unsupervised anomaly detection algorithms often suffer from excessive false alarms, especially in the presence of concept drifts resulting from network re-configurations or maintenance. To tackle this challenge and improve the overall performance of unsupervised anomaly detection algorithms, we propose to use active learning to introduce and benefit from the feedback of operators, who can verify the alarms (both false and true ones) and label the corresponding KPIs with reasonable effort. Specifically, we develop three query strategies to select the most informative and representative samples to label. We also develop an efficient method to update the weights of Isolation Forest and optimally adjust the decision threshold, so as to eventually improve the performance of detection model. The experiments with one public dataset and one proprietary dataset demonstrate that our active learning empowered anomaly detection pipeline could achieve performance gain, in terms of F1-score, more than 50% over the baseline algorithm. It also outperforms the existing active learning based methods by approximately 6%-10%, with significantly reduced budget (the ratio of samples to be labeled)., Comment: Paper accepted at the AIOPS 2021 workshop from the 19th International Conference on Service-Oriented Computing (ICSOC). (https://aiops2021.github.io/accepted_papers/index.html)

We introduce a high-quality and large-scale Vietnamese-English parallel dataset of 3.02M sentence pairs, which is 2.9M pairs larger than the benchmark Vietnamese-English machine translation corpus IWSLT15. We conduct experiments comparing strong neural baselines and well-known automatic translation engines on our dataset and find that in both automatic and human evaluations: the best performance is obtained by fine-tuning the pre-trained sequence-to-sequence denoising auto-encoder mBART. To our best knowledge, this is the first large-scale Vietnamese-English machine translation study. We hope our publicly available dataset and study can serve as a starting point for future research and applications on Vietnamese-English machine translation., Comment: To appear in Proceedings of EMNLP 2021 (main conference). The first three authors contribute equally to this work

Biology offers many examples of large-scale, complex, concurrent systems: many processes take place in parallel, compete on resources and influence each other's behavior. The scalable modeling of biological systems continues to be a very active field of research. In this paper we introduce a new approach based on Event-B, a state-based formal method with refinement as its central ingredient, allowing us to check for model consistency step-by-step in an automated way. Our approach based on functions leads to an elegant and concise modeling method. We demonstrate this approach by constructing what is, to our knowledge, the largest ever built Event-B model, describing the ErbB signaling pathway, a key evolutionary pathway with a significant role in development and in many types of cancer. The Event-B model for the ErbB pathway describes 1320 molecular reactions through 242 events.

As of 2020 when the COVID-19 pandemic is full-blown on a global scale, people's need to have access to legitimate information regarding COVID-19 is more urgent than ever, especially via online media where the abundance of irrelevant information overshadows the more informative ones. In response to such, we proposed a model that, given an English tweet, automatically identifies whether that tweet bears informative content regarding COVID-19 or not. By ensembling different BERTweet model configurations, we have achieved competitive results that are only shy of those by top performing teams by roughly 1% in terms of F1 score on the informative class. In the post-competition period, we have also experimented with various other approaches that potentially boost generalization to a new dataset.

The Nonlinear Grey Bernoulli Model NGBM(1, 1) is a recently developed grey model which has various applications in different fields, mainly due to its accuracy in handling small time-series datasets with nonlinear variations. In this paper, to fully improve the accuracy of this model, a novel model is proposed, namely Rolling Optimized Nonlinear Grey Bernoulli Model RONGBM(1, 1). This model combines the rolling mechanism with the simultaneous optimization of all model parameters (exponential, background value and initial condition). The accuracy of this new model has significantly been proven through forecasting Vietnam's GDP from 2013 to 2018, before it is applied to predict the total COVID-19 infected cases globally by day., Comment: Accepted paper at the 2020 International Congress of Grey Systems and Uncertainty Analysis (GSUA)

We present experimental studies of atom-ion collisions using buffer-gas cooled, trapped ytterbium (Yb$^+$) ions immersed in potassium (K) vapor. The range of the collisional temperature is on the order of several hundred kelvin (thermal regime). We have determined various collisional rate coefficients of the Yb$^+$ ion per K-atom number density. We find the upper bounds of charge-exchange rate coefficients $\kappa_{\rm ce}$ to be $(12.7\pm1.6)\times10^{-14}$ cm$^3$s$^{-1}$ for K-$^{171}$Yb$^+$ and $(5.3\pm0.7)\times10^{-14}$ cm$^3$s$^{-1}$ for K-$^{172}$Yb$^+$. For both isotopes, the spin-destruction rate coefficient $\kappa_{\rm sd}$ has an upper bound at $(1.46\pm0.77)\times10^{-9}$ cm$^3$s$^{-1}$. The spin-exchange rate coefficient $\kappa_{\rm se}$ is measured to be $(1.64\pm0.51)\times10^{-9}$ cm$^3$s$^{-1}$. The relatively low charge-exchange rate reported here demonstrates the advantage of using K atoms to sympathetically cool Yb$^+$ ions, and the relatively high spin-exchange rate may benefit research work in quantum metrology and quantum information processing on an atom-ion platform using K atoms and Yb$^+$ ions., Comment: 6 pages, 4 figures

Hindered adsorbed layers completely filling the nanopores must cause significant deviations from the classical BET isotherms for multimolecular adsorption of vapor in porous solids. Since the point of transition from free to hindered adsorption moves into narrower nanopores as the pore vapor pressure (or humidity) decreases, the area of the surface of multimolecular adsorption layer exposed to vapor gets modified by an area reduction factor that decreases with increasing thickness of the layer, and thus also with increasing vapor pressure. The area reduction factor does not affect the rates of direct adsorption or condensation from the vapor, which represent a local process, but imposes a lateral constraint on the total area and volume of the free portion of the adsorption layer in direct contact with vapor. Assuming a power-law for the dependence of the area reduction factor on the number of molecular layers, one can express the modified isotherm in terms of logarithmic or polylogarithmic (a.k.a Jonqui\`ere) functions. Comparisons with some published experimentally obtained isotherms show that the present modification of the BET theory for hindered adsorption goes in the right direction. Detailed calibration of the theory and an extension for indirect communication of vapor molecules with the molecules adsorbed in nanopores less than a few nm wide will require further research., Comment: 15 pages, 4 figures