Publications (273)
BioDiscoveryAgent: An AI Agent for Designing Genetic Perturbation Experiments
Yusuf Roohani, Andrew Lee, Qian Huang +6
Agents based on large language models have shown great potential in accelerating scientific discovery by leveraging their rich background knowledge and reasoning capabilities. In t…
Unifying Human and Statistical Evaluation for Natural Language Generation
Tatsunori B. Hashimoto, Hugh Zhang, Percy Liang
How can we measure whether a natural language generation system produces both high quality and diverse outputs? Human evaluation captures quality but not diversity, as it does not…
The Statistics of Streaming Sparse Regression
Jacob Steinhardt, Stefan Wager, Percy Liang
We present a sparse analogue to stochastic gradient descent that is guaranteed to perform well under similar conditions to the lasso. In the linear regression setup with irrepresen…
The Mighty ToRR: A Benchmark for Table Reasoning and Robustness
Shir Ashury-Tahan, Yifan Mai, Rajmohan C +8
Despite its real-world significance, model performance on tabular data remains underexplored, leaving uncertainty about which model to rely on and which prompt configuration to ado…
The California Report on Frontier AI Policy
Rishi Bommasani, Scott R. Singer, Ruth E. Appel +20
The innovations emerging at the frontier of artificial intelligence (AI) are poised to create historic opportunities for humanity but also raise complex policy challenges. Continue…
Length-Controlled AlpacaEval: A Simple Way to Debias Automatic Evaluators
Yann Dubois, Balázs Galambosi, Percy Liang +1
LLM-based auto-annotators have become a key component of the LLM development process due to their cost-effectiveness and scalability compared to human-based evaluation. However, th…
Language Models May Verbatim Complete Text They Were Not Explicitly Trained On
Ken Ziyu Liu, Christopher A. Choquette-Choo, Matthew Jagielski +4
An important question today is whether a given text was used to train a large language model (LLM). A \emph{completion} test is often employed: check if the LLM completes a suffici…
OpenVLA: An Open-Source Vision-Language-Action Model
Moo Jin Kim, Karl Pertsch, Siddharth Karamcheti +15
Large policies pretrained on a combination of Internet-scale vision-language data and diverse robot demonstrations have the potential to change how we teach robots new skills: rath…
RoboArena: Distributed Real-World Evaluation of Generalist Robot Policies
Pranav Atreya, Karl Pertsch, Tony Lee +29
Comprehensive, unbiased, and comparable evaluation of modern generalist policies is uniquely challenging: existing approaches for robot benchmarking typically rely on heavy standar…
Stronger Data Poisoning Attacks Break Data Sanitization Defenses
Pang Wei Koh, Jacob Steinhardt, Percy Liang
Machine learning models trained on data from the outside world can be corrupted by data poisoning attacks that inject malicious points into the models' training sets. A common defe…
Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization
Shiori Sagawa, Pang Wei Koh, Tatsunori B. Hashimoto +1
Overparameterized neural networks can be highly accurate on average on an i.i.d. test set yet consistently fail on atypical groups of the data (e.g., by learning spurious correlati…
Anticipatory Music Transformer
John Thickstun, David Hall, Chris Donahue +1
We introduce anticipation: a method for constructing a controllable generative model of a temporal point process (the event process) conditioned asynchronously on realizations of a…
Delete, Retrieve, Generate: A Simple Approach to Sentiment and Style Transfer
Juncen Li, Robin Jia, He He +1
We consider the task of text attribute transfer: transforming a sentence to alter a specific attribute (e.g., sentiment) while preserving its attribute-independent content (e.g., c…
Ecosystem Graphs: The Social Footprint of Foundation Models
Rishi Bommasani, Dilara Soylu, Thomas I. Liao +2
Foundation models (e.g. ChatGPT, StableDiffusion) pervasively influence society, warranting immediate social attention. While the models themselves garner much attention, to accura…
Concept Bottleneck Models
Pang Wei Koh, Thao Nguyen, Yew Siang Tang +4
We seek to learn models that we can interact with using high-level concepts: if the model did not think there was a bone spur in the x-ray, would it still predict severe arthritis?…
AlpacaFarm: A Simulation Framework for Methods that Learn from Human Feedback
Yann Dubois, Xuechen Li, Rohan Taori +6
Large language models (LLMs) such as ChatGPT have seen widespread adoption due to their strong instruction-following abilities. Developing these LLMs involves a complex yet poorly…
Improving Dynamic Object Interactions in Text-to-Video Generation with AI Feedback
Hiroki Furuta, Heiga Zen, Dale Schuurmans +4
Large text-to-video models hold immense potential for a wide range of downstream applications. However, they struggle to accurately depict dynamic object interactions, often result…
Reinforcement Learning for Machine Learning Engineering Agents
Sherry Yang, Joy He-Yueya, Percy Liang
Existing agents for solving tasks such as ML engineering rely on prompting powerful language models. As a result, these agents do not improve with more experience. In this paper, w…
MLE-Smith: Scaling MLE Tasks with Automated Multi-Agent Pipeline
Rushi Qiang, Yuchen Zhuang, Anikait Singh +4
While Language Models (LMs) have made significant progress in automating machine learning engineering (MLE), the acquisition of high-quality MLE training data is significantly cons…
Learning Language Games through Interaction
Sida I. Wang, Percy Liang, Christopher D. Manning
We introduce a new language learning setting relevant to building adaptive natural language interfaces. It is inspired by Wittgenstein's language games: a human wishes to accomplis…
One-sided Matrix Completion from Two Observations Per Row
Steven Cao, Percy Liang, Gregory Valiant
Given only a few observed entries from a low-rank matrix , matrix completion is the problem of imputing the missing entries, and it formalizes a wide range of real-world setting…
CoAuthor: Designing a Human-AI Collaborative Writing Dataset for Exploring Language Model Capabilities
Mina Lee, Percy Liang, Qian Yang
Large language models (LMs) offer unprecedented language generation capabilities and exciting opportunities for interaction design. However, their highly context-dependent capabili…
The 2025 Foundation Model Transparency Index
Alexander Wan, Kevin Klyman, Sayash Kapoor +5
Foundation model developers are among the world's most important companies. As these companies become increasingly consequential, how do their transparency practices evolve? The 20…
Robust Distortion-free Watermarks for Language Models
Rohith Kuditipudi, John Thickstun, Tatsunori Hashimoto +1
We propose a methodology for planting watermarks in text from an autoregressive language model that are robust to perturbations without changing the distribution over text up to a…
Break-It-Fix-It: Unsupervised Learning for Program Repair
Michihiro Yasunaga, Percy Liang
We consider repair tasks: given a critic (e.g., compiler) that assesses the quality of an input, the goal is to train a fixer that converts a bad example (e.g., code with syntax er…
The Responsible Foundation Model Development Cheatsheet: A Review of Tools & Resources
Shayne Longpre, Stella Biderman, Alon Albalak +20
Foundation model development attracts a rapidly expanding body of contributors, scientists, and applications. To help shape responsible development practices, we introduce the Foun…
On the Learnability of Watermarks for Language Models
Chenchen Gu, Xiang Lisa Li, Percy Liang +1
Watermarking of language model outputs enables statistical detection of model-generated text, which can mitigate harms and misuses of language models. Existing watermarking strateg…
Conditional probing: measuring usable information beyond a baseline
John Hewitt, Kawin Ethayarajh, Percy Liang +1
Probing experiments investigate the extent to which neural representations make properties -- like part-of-speech -- predictable. One suggests that a representation encodes a prope…
The price of debiasing automatic metrics in natural language evaluation
Arun Tejasvi Chaganty, Stephen Mussman, Percy Liang
For evaluating generation systems, automatic metrics such as BLEU cost nothing to run but have been shown to correlate poorly with human judgment, leading to systematic bias agains…
FlexGen: High-Throughput Generative Inference of Large Language Models with a Single GPU
Ying Sheng, Lianmin Zheng, Binhang Yuan +11
The high computational and memory requirements of large language model (LLM) inference make it feasible only with multiple high-end accelerators. Motivated by the emerging demand f…
The EOS Decision and Length Extrapolation
Benjamin Newman, John Hewitt, Percy Liang +1
Extrapolation to unseen sequence lengths is a challenge for neural generative models of language. In this work, we characterize the effect on length extrapolation of a modeling dec…
A Hitting Time Analysis of Stochastic Gradient Langevin Dynamics
Yuchen Zhang, Percy Liang, Moses Charikar
We study the Stochastic Gradient Langevin Dynamics (SGLD) algorithm for non-convex optimization. The algorithm performs stochastic gradient descent, where in each step it injects a…
Cheaply Evaluating Inference Efficiency Metrics for Autoregressive Transformer APIs
Deepak Narayanan, Keshav Santhanam, Peter Henderson +3
Large language models (LLMs) power many state-of-the-art systems in natural language processing. However, these models are extremely computationally expensive, even at inference ti…
Just Train Twice: Improving Group Robustness without Training Group Information
Evan Zheran Liu, Behzad Haghgoo, Annie S. Chen +5
Standard training via empirical risk minimization (ERM) can produce models that achieve high accuracy on average but low accuracy on certain groups, especially in the presence of s…
Reified Context Models
Jacob Steinhardt, Percy Liang
A classic tension exists between exact inference in a simple model and approximate inference in a complex model. The latter offers expressivity and thus accuracy, but the former pr…
International Scientific Report on the Safety of Advanced AI (Interim Report)
Yoshua Bengio, Sören Mindermann, Daniel Privitera +41
This is the interim publication of the first International Scientific Report on the Safety of Advanced AI. The report synthesises the scientific understanding of general-purpose AI…
WorldGym: World Model as An Environment for Policy Evaluation
Julian Quevedo, Ansh Kumar Sharma, Yixiang Sun +3
Evaluating robot control policies is difficult: real-world testing is costly, and handcrafted simulators require manual effort to improve in realism and generality. We propose a wo…
Fantastic Pretraining Optimizers and Where to Find Them
Kaiyue Wen, David Hall, Tengyu Ma +1
AdamW has long been the dominant optimizer in language model pretraining, despite numerous claims that alternative optimizers offer 1.4 to 2x speedup. We posit that two methodologi…
Deep Bidirectional Language-Knowledge Graph Pretraining
Michihiro Yasunaga, Antoine Bosselut, Hongyu Ren +4
Pretraining a language model (LM) on text has been shown to help various downstream NLP tasks. Recent works show that a knowledge graph (KG) can complement text data, offering stru…
Model Equality Testing: Which Model Is This API Serving?
Irena Gao, Percy Liang, Carlos Guestrin
Users often interact with large language models through black-box inference APIs, both for closed- and open-weight models (e.g., Llama models are popularly accessed via Amazon Bedr…
Generative Agents: Interactive Simulacra of Human Behavior
Joon Sung Park, Joseph C. O'Brien, Carrie J. Cai +3
Believable proxies of human behavior can empower interactive applications ranging from immersive environments to rehearsal spaces for interpersonal communication to prototyping too…
Composed Fine-Tuning: Freezing Pre-Trained Denoising Autoencoders for Improved Generalization
Sang Michael Xie, Tengyu Ma, Percy Liang
We focus on prediction problems with structured outputs that are subject to output validity constraints, e.g. pseudocode-to-code translation where the code must compile. While labe…
Learning Symmetric Collaborative Dialogue Agents with Dynamic Knowledge Graph Embeddings
He He, Anusha Balakrishnan, Mihail Eric +1
We study a symmetric collaborative dialogue setting in which two agents, each with private knowledge, must strategically communicate to achieve a common goal. The open-ended dialog…
SPoC: Search-based Pseudocode to Code
Sumith Kulal, Panupong Pasupat, Kartik Chandra +4
We consider the task of mapping pseudocode to long programs that are functionally correct. Given test cases as a mechanism to validate programs, we search over the space of possibl…
Ecosystem-level Analysis of Deployed Machine Learning Reveals Homogeneous Outcomes
Connor Toups, Rishi Bommasani, Kathleen A. Creel +3
Machine learning is traditionally studied at the model level: researchers measure and improve the accuracy, robustness, bias, efficiency, and other dimensions of specific models. I…
MLE-Dojo: Interactive Environments for Empowering LLM Agents in Machine Learning Engineering
Rushi Qiang, Yuchen Zhuang, Yinghao Li +8
We introduce MLE-Dojo, a Gym-style framework for systematically reinforcement learning, evaluating, and improving autonomous large language model (LLM) agents in iterative machine…
AI Risk Categorization Decoded (AIR 2024): From Government Regulations to Corporate Policies
Yi Zeng, Kevin Klyman, Andy Zhou +6
We present a comprehensive AI risk taxonomy derived from eight government policies from the European Union, United States, and China and 16 company policies worldwide, making a sig…
VLAW: Iterative Co-Improvement of Vision-Language-Action Policy and World Model
Yanjiang Guo, Tony Lee, Lucy Xiaoyang Shi +3
The goal of this paper is to improve the performance and reliability of vision-language-action (VLA) models through iterative online interaction. Since collecting policy rollouts i…
AIR-Bench 2024: A Safety Benchmark Based on Risk Categories from Regulations and Policies
Yi Zeng, Yu Yang, Andy Zhou +9
Foundation models (FMs) provide societal benefits but also amplify risks. Governments, companies, and researchers have proposed regulatory frameworks, acceptable use policies, and…
Establishing Best Practices for Building Rigorous Agentic Benchmarks
Yuxuan Zhu, Tengjun Jin, Yada Pruksachatkun +22
Benchmarks are essential for quantitatively tracking progress in AI. As AI agents become increasingly capable, researchers and practitioners have introduced agentic benchmarks to e…
AHELM: A Holistic Evaluation of Audio-Language Models
Tony Lee, Haoqin Tu, Chi Heem Wong +6
Evaluations of audio-language models (ALMs) -- multimodal models that take interleaved audio and text as input and output text -- are hindered by the lack of standardized benchmark…
LM-Critic: Language Models for Unsupervised Grammatical Error Correction
Michihiro Yasunaga, Jure Leskovec, Percy Liang
Training a model for grammatical error correction (GEC) requires a set of labeled ungrammatical / grammatical sentence pairs, but manually annotating such pairs can be expensive. R…
Self-Verified Distillation: Your Language Model Is Secretly Its Own Synthetic Data Pipeline
Tony Lee, Percy Liang
Can post-trained large language models (LLMs) further improve themselves using only unlabeled prompts, without external teachers or feedback from tools? We study this setting start…
Picking on the Same Person: Does Algorithmic Monoculture lead to Outcome Homogenization?
Rishi Bommasani, Kathleen A. Creel, Ananya Kumar +2
As the scope of machine learning broadens, we observe a recurring theme of algorithmic monoculture: the same systems, or systems that share components (e.g. training data), are dep…
Estimating Mixture Models via Mixtures of Polynomials
Sida I. Wang, Arun Tejasvi Chaganty, Percy Liang
Mixture modeling is a general technique for making any simple model more expressive through weighted combination. This generality and simplicity in part explains the success of the…
Extracting books from production language models
Ahmed Ahmed, A. Feder Cooper, Sanmi Koyejo +1
Many unresolved legal questions over LLMs and copyright center on memorization: whether specific training data have been encoded in the model's weights during training, and whether…
Surgical Fine-Tuning Improves Adaptation to Distribution Shifts
Yoonho Lee, Annie S. Chen, Fahim Tajwar +4
A common approach to transfer learning under distribution shift is to fine-tune the last few layers of a pre-trained model, preserving learned features while also adapting to the n…
Understanding and Mitigating the Tradeoff Between Robustness and Accuracy
Aditi Raghunathan, Sang Michael Xie, Fanny Yang +2
Adversarial training augments the training set with perturbations to improve the robust error (over worst-case perturbations), but it often leads to an increase in the standard err…
LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individuals
Joon Sung Park, Carolyn Q. Zou, Jonne Kamphorst +8
Machine learning can predict human behavior well when substantial structured data are available for well-defined outcomes. Such models are typically outcome-specific, however, requ…
MedHELM: Holistic Evaluation of Large Language Models for Medical Tasks
Suhana Bedi, Hejie Cui, Miguel Fuentes +78
While large language models (LLMs) achieve near-perfect scores on medical licensing exams, these evaluations inadequately reflect the complexity and diversity of real-world clinica…
Trustworthy Social Bias Measurement
Rishi Bommasani, Percy Liang
How do we design measures of social bias that we trust? While prior work has introduced several measures, no measure has gained widespread trust: instead, mounting evidence argues…
Selective Question Answering under Domain Shift
Amita Kamath, Robin Jia, Percy Liang
To avoid giving wrong answers, question answering (QA) models need to know when to abstain from answering. Moreover, users often ask questions that diverge from the model's trainin…
Calibrated ensembles can mitigate accuracy tradeoffs under distribution shift
Ananya Kumar, Tengyu Ma, Percy Liang +1
We often see undesirable tradeoffs in robust machine learning where out-of-distribution (OOD) accuracy is at odds with in-distribution (ID) accuracy: a robust classifier obtained v…
Instruction Following without Instruction Tuning
John Hewitt, Nelson F. Liu, Percy Liang +1
Instruction tuning commonly means finetuning a language model on instruction-response pairs. We discover two forms of adaptation (tuning) that are deficient compared to instruction…
Foundation Model Transparency Reports
Rishi Bommasani, Kevin Klyman, Shayne Longpre +5
Foundation models are critical digital technologies with sweeping societal impact that necessitates transparency. To codify how foundation model developers should provide transpare…
Beyond I.I.D.: Three Levels of Generalization for Question Answering on Knowledge Bases
Yu Gu, Sue Kase, Michelle Vanni +4
Existing studies on question answering on knowledge bases (KBQA) mainly operate with the standard i.i.d assumption, i.e., training distribution over questions is the same as the te…
Blackbox Model Provenance via Palimpsestic Membership Inference
Rohith Kuditipudi, Jing Huang, Sally Zhu +3
Suppose Alice trains an open-weight language model and Bob uses a blackbox derivative of Alice's model to produce text. Can Alice prove that Bob is using her model, either by query…
MedAlign: A Clinician-Generated Dataset for Instruction Following with Electronic Medical Records
Scott L. Fleming, Alejandro Lozano, William J. Haberkorn +27
The ability of large language models (LLMs) to follow natural language instructions with human-level fluency suggests many opportunities in healthcare to reduce administrative burd…
Llamas Know What GPTs Don't Show: Surrogate Models for Confidence Estimation
Vaishnavi Shrivastava, Percy Liang, Ananya Kumar
To maintain user trust, large language models (LLMs) should signal low confidence on examples where they are incorrect, instead of misleading the user. The standard approach of est…
Prefix-Tuning: Optimizing Continuous Prompts for Generation
Xiang Lisa Li, Percy Liang
Fine-tuning is the de facto way to leverage large pretrained language models to perform downstream tasks. However, it modifies all the language model parameters and therefore neces…
PRODIGY: Enabling In-context Learning Over Graphs
Qian Huang, Hongyu Ren, Peng Chen +4
In-context learning is the ability of a pretrained model to adapt to novel and diverse downstream tasks by conditioning on prompt examples, without optimizing any parameters. While…
Diffusion-LM Improves Controllable Text Generation
Xiang Lisa Li, John Thickstun, Ishaan Gulrajani +2
Controlling the behavior of language models (LMs) without re-training is a major open problem in natural language generation. While recent works have demonstrated successes on cont…
Learning Autocomplete Systems as a Communication Game
Mina Lee, Tatsunori B. Hashimoto, Percy Liang
We study textual autocomplete---the task of predicting a full sentence from a partial sentence---as a human-machine communication game. Specifically, we consider three competing go…
Backpack Language Models
John Hewitt, John Thickstun, Christopher D. Manning +1
We present Backpacks: a new neural architecture that marries strong modeling performance with an interface for interpretability and control. Backpacks learn multiple non-contextual…
Learning Executable Semantic Parsers for Natural Language Understanding
Percy Liang
For building question answering systems and natural language interfaces, semantic parsing has emerged as an important and powerful paradigm. Semantic parsers map natural language i…
Large Language Models Can Be Strong Differentially Private Learners
Xuechen Li, Florian Tramèr, Percy Liang +1
Differentially Private (DP) learning has seen limited success for building large deep learning models of text, and straightforward attempts at applying Differentially Private Stoch…
RoboReward: General-Purpose Vision-Language Reward Models for Robotics
Tony Lee, Andrew Wagenmaker, Karl Pertsch +3
A well-designed reward is critical for effective reinforcement learning-based policy improvement. In real-world robotics, obtaining such rewards typically requires either labor-int…
International AI Safety Report
Yoshua Bengio, Sören Mindermann, Daniel Privitera +93
The first International AI Safety Report comprehensively synthesizes the current evidence on the capabilities, risks, and safety of advanced AI systems. The report was mandated by…
Evaluating Human-Language Model Interaction
Mina Lee, Megha Srivastava, Amelia Hardy +15
Many real-world applications of language models (LMs), such as writing assistance and code autocomplete, involve human-LM interaction. However, most benchmarks are non-interactive…
VHELM: A Holistic Evaluation of Vision Language Models
Tony Lee, Haoqin Tu, Chi Heem Wong +8
Current benchmarks for assessing vision-language models (VLMs) often focus on their perception or problem-solving capabilities and neglect other critical aspects such as fairness,…
Robustness to Spurious Correlations via Human Annotations
Megha Srivastava, Tatsunori Hashimoto, Percy Liang
The reliability of machine learning systems critically assumes that the associations between features and labels remain similar between training and test distributions. However, un…
Uncertainty Sampling is Preconditioned Stochastic Gradient Descent on Zero-One Loss
Stephen Mussmann, Percy Liang
Uncertainty sampling, a popular active learning algorithm, is used to reduce the amount of data required to learn a classifier, but it has been observed in practice to converge to…
Removing Spurious Features can Hurt Accuracy and Affect Groups Disproportionately
Fereshte Khani, Percy Liang
The presence of spurious features interferes with the goal of obtaining robust models that perform well across many groups within the population. A natural remedy is to remove spur…
Understanding Black-box Predictions via Influence Functions
Pang Wei Koh, Percy Liang
How can we explain the predictions of a black-box model? In this paper, we use influence functions -- a classic technique from robust statistics -- to trace a model's prediction th…
In-N-Out: Pre-Training and Self-Training using Auxiliary Information for Out-of-Distribution Robustness
Sang Michael Xie, Ananya Kumar, Robbie Jones +3
Consider a prediction setting with few in-distribution labeled examples and many unlabeled examples both in- and out-of-distribution (OOD). The goal is to learn a model which perfo…
Mapping Natural Language Commands to Web Elements
Panupong Pasupat, Tian-Shun Jiang, Evan Zheran Liu +2
The web provides a rich, open-domain environment with textual, structural, and spatial properties. We propose a new task for grounding language in this environment: given a natural…
Beat the long tail: Distribution-Aware Speculative Decoding for RL Training
Zelei Shao, Vikranth Srivatsa, Sanjana Srivastava +12
Reinforcement learning(RL) post-training has become essential for aligning large language models (LLMs), yet its efficiency is increasingly constrained by the rollout phase, where…
Certified Robustness to Adversarial Word Substitutions
Robin Jia, Aditi Raghunathan, Kerem Göksel +1
State-of-the-art NLP models can often be fooled by adversaries that apply seemingly innocuous label-preserving transformations (e.g., paraphrasing) to input text. The number of pos…
MLAgentBench: Evaluating Language Agents on Machine Learning Experimentation
Qian Huang, Jian Vora, Percy Liang +1
A central aspect of machine learning research is experimentation, the process of designing and running experiments, analyzing the results, and iterating towards some positive outco…
Is a Caption Worth a Thousand Images? A Controlled Study for Representation Learning
Shibani Santurkar, Yann Dubois, Rohan Taori +2
The development of CLIP [Radford et al., 2021] has sparked a debate on whether language supervision can result in vision models with more transferable representations than traditio…
Data-efficient pre-training by scaling synthetic megadocs
Konwoo Kim, Suhas Kotha, Yejin Choi +3
Synthetic data augmentation has emerged as a promising solution when pre-training is constrained by data rather than compute. We study how to design synthetic data algorithms that…
Extracting memorized pieces of (copyrighted) books from open-weight language models
A. Feder Cooper, Mark A. Lemley, Allison Casasola +6
Plaintiffs and defendants in copyright lawsuits over generative AI often make sweeping, opposing claims about the extent to which large language models (LLMs) memorize protected ex…
Do Question Answering Modeling Improvements Hold Across Benchmarks?
Nelson F. Liu, Tony Lee, Robin Jia +1
Do question answering (QA) modeling improvements (e.g., choice of architecture and training procedure) hold consistently across the diverse landscape of QA benchmarks? To study thi…
Swords: A Benchmark for Lexical Substitution with Improved Data Coverage and Quality
Mina Lee, Chris Donahue, Robin Jia +2
We release a new benchmark for lexical substitution, the task of finding appropriate substitutes for a target word in a context. To assist humans with writing, lexical substitution…
Simultaneous diagonalization: the asymmetric, low-rank, and noisy settings
Volodymyr Kuleshov, Arun Tesjavi Chaganty, Percy Liang
Simultaneous matrix diagonalization is used as a subroutine in many machine learning problems, including blind source separation and paramater estimation in latent variable models.…
Introducing v0.5 of the AI Safety Benchmark from MLCommons
Bertie Vidgen, Adarsh Agrawal, Ahmed M. Ahmed +97
This paper introduces v0.5 of the AI Safety Benchmark, which has been created by the MLCommons AI Safety Working Group. The AI Safety Benchmark has been designed to assess the safe…
SQuAD: 100,000+ Questions for Machine Comprehension of Text
Pranav Rajpurkar, Jian Zhang, Konstantin Lopyrev +1
We present the Stanford Question Answering Dataset (SQuAD), a new reading comprehension dataset consisting of 100,000+ questions posed by crowdworkers on a set of Wikipedia article…
Robust Encodings: A Framework for Combating Adversarial Typos
Erik Jones, Robin Jia, Aditi Raghunathan +1
Despite excellent performance on many tasks, NLP systems are easily fooled by small adversarial perturbations of inputs. Existing procedures to defend against such perturbations ar…
Machine Unlearning Doesn't Do What You Think: Lessons for Generative AI Policy and Research
A. Feder Cooper, Christopher A. Choquette-Choo, Miranda Bogen +34
"Machine unlearning" is a popular proposed solution for mitigating the existence of content in an AI model that is problematic for legal or moral reasons, including privacy, copyri…
Auditing Prompt Caching in Language Model APIs
Chenchen Gu, Xiang Lisa Li, Rohith Kuditipudi +2
Prompt caching in large language models (LLMs) results in data-dependent timing variations: cached prompts are processed faster than non-cached prompts. These timing differences in…