research
∙
09/14/2023
Safety-Tuned LLaMAs: Lessons From Improving the Safety of Large Language Models that Follow Instructions
Training large language models to follow instructions makes them perform...
share
research
∙
08/28/2023
Identifying and Mitigating the Security Risks of Generative AI
Every major technical invention resurfaces the dual-use dilemma – the ne...
research
∙
08/17/2023
Accelerating Aggregation Queries on Unstructured Streams of Data
Analysts and scientists are interested in querying streams of video, aud...
research
∙
08/09/2023
Where's the Liability in Harmful AI Speech?
Generative AI, in particular text-based "foundation models" (large model...
research
∙
07/28/2023
Robust Distortion-free Watermarks for Language Models
We propose a methodology for planting watermarks in text from an autoreg...
research
∙
03/30/2023
Whose Opinions Do Language Models Reflect?
Language models (LMs) are increasingly being used in open-ended contexts...
research
∙
03/28/2023
Foundation Models and Fair Use
Existing foundation models are trained on copyrighted material. Deployin...
research
∙
02/26/2023
Navigating the Grey Area: Expressions of Overconfidence and Uncertainty in Language Models
Despite increasingly fluent, relevant, and coherent language generation,...
research
∙
02/23/2023
Out-of-Domain Robustness via Targeted Augmentations
Models trained on one set of domains often suffer performance drops on u...
research
∙
02/11/2023
Exploiting Programmatic Behavior of LLMs: Dual-Use Through Standard Security Attacks
Recent advances in instruction-following large language models (LLMs) ha...
research
∙
12/21/2022
Tracing and Removing Data Errors in Natural Language Generation Datasets
Recent work has identified noisy and misannotated data as a core cause o...
research
∙
12/20/2022
Privacy-Preserving Domain Adaptation of Semantic Parsers
Task-oriented dialogue systems often assist users with personal or confi...
research
∙
11/16/2022
Holistic Evaluation of Language Models
Language models (LMs) are becoming the foundation for almost all major l...
research
∙
11/09/2022
ZK-IMG: Attested Images via Zero-Knowledge Proofs to Fight Disinformation
Over the past few years, AI methods of generating images have been incre...
research
∙
11/07/2022
Easily Accessible Text-to-Image Generation Amplifies Demographic Stereotypes at Large Scale
Machine learning models are now able to convert user-written text descri...
research
∙
10/27/2022
Contrastive Decoding: Open-ended Text Generation as Optimization
Likelihood, although useful as a training loss, is a poor search objecti...
research
∙
10/17/2022
Scaling up Trustless DNN Inference with Zero-Knowledge Proofs
As ML models have increased in capabilities and accuracy, so has the com...
research
∙
10/15/2022
A Closer Look at the Calibration of Differentially Private Learners
We systematically study the calibration of classifiers trained with diff...
research
∙
07/01/2022
When Does Differentially Private Learning Not Suffer in High Dimensions?
Large pretrained models can be privately fine-tuned to achieve performan...
research
∙
05/26/2022
Undersampling is a Minimax Optimal Robustness Intervention in Nonparametric Classification
While a broad range of techniques have been proposed to tackle distribut...
research
∙
04/21/2022
Spurious Correlations in Reference-Free Evaluation of Text Generation
Model-based, reference-free evaluation metrics have been proposed as a f...
research
∙
04/13/2022
Distributionally Robust Models with Parametric Likelihood Ratios
As machine learning models are deployed ever more broadly, it becomes in...
research
∙
03/21/2022
Language modeling via stochastic processes
Modern language models can generate high-quality short texts. However, t...
research
∙
02/07/2022
Jury Learning: Integrating Dissenting Voices into Machine Learning Models
Whose labels should a machine learning (ML) algorithm learn to emulate? ...
research
∙
12/24/2021
Is Importance Weighting Incompatible with Interpolating Classifiers?
Importance weighting is a classic technique to handle distribution shift...
research
∙
12/09/2021
Extending the WILDS Benchmark for Unsupervised Adaptation
Machine learning systems deployed in the wild are often trained on a sou...
research
∙
10/12/2021
Large Language Models Can Be Strong Differentially Private Learners
Differentially Private (DP) learning has seen limited success for buildi...
research
∙
06/07/2021
Measuring Conversational Uptake: A Case Study on Student-Teacher Interactions
In conversation, uptake happens when a speaker builds on the contributio...
research
∙
04/12/2021
On the Inductive Bias of Masked Language Modeling: From Statistical to Syntactic Dependencies
We study how masking and predicting tokens in an unsupervised fashion ca...
research
∙
03/18/2021
Modeling the Second Player in Distributionally Robust Optimization
Distributionally robust optimization (DRO) provides a framework for trai...
research
∙
02/02/2021
The GEM Benchmark: Natural Language Generation, its Evaluation and Metrics
We introduce GEM, a living benchmark for natural language Generation (NL...
research
∙
09/09/2020
Task-agnostic Indexes for Deep Learning-based Queries over Unstructured Data
Unstructured data is now commonly queried by using target deep neural ne...
research
∙
07/28/2020
Distributionally Robust Losses for Latent Covariate Mixtures
While modern large-scale datasets often consist of heterogeneous subpopu...
research
∙
07/13/2020
Robustness to Spurious Correlations via Human Annotations
The reliability of machine learning systems critically assumes that the ...
research
∙
04/30/2020
Improved Natural Language Generation via Loss Truncation
Neural language models are usually trained to match the distributional p...
research
∙
04/02/2020
Approximate Selection with Guarantees using Proxies
Due to the falling costs of data acquisition and storage, researchers an...
research
∙
07/12/2018