activity
20242026
collaborators

7 papers

cs.LG2026

When Should We Introduce Safety Interventions During Pretraining?

Dylan Sam, Sachin Goyal, Pratyush Maini +2

Prior work has shown that safety interventions applied during pretraining, such as removing and rephrasing harmful content, can substantially improve the robustness of the resultin…

cs.LG2025

Predicting the Performance of Black-box LLMs through Follow-up Queries

Dylan Sam, Marc Finzi, J. Zico Kolter

Reliably predicting the behavior of language models -- such as whether their outputs are correct or have been adversarially manipulated -- is a fundamentally challenging task. This…

cs.LG2025

Analyzing Similarity Metrics for Data Selection for Language Model Pretraining

Dylan Sam, Ayan Chakrabarti, Afshin Rostamizadeh +3

Measuring similarity between training examples is critical for curating high-quality and diverse pretraining datasets for language models. However, similarity is typically computed…

cs.LG2025

Safety Pretraining: Toward the Next Generation of Safe AI

Pratyush Maini, Sachin Goyal, Dylan Sam +7

As large language models (LLMs) are increasingly deployed in high-stakes settings, the risk of generating harmful or toxic content remains a central challenge. Post-hoc alignment m…

cs.LG2025

Evaluating Language Model Reasoning about Confidential Information

Dylan Sam, Alexander Robey, Andy Zou +2

As language models are increasingly deployed as autonomous agents in high-stakes settings, ensuring that they reliably follow user-defined rules has become a critical safety concer…

cs.LG2025

Finetuning CLIP to Reason about Pairwise Differences

Dylan Sam, Devin Willmott, Joao D. Semedo +1

Vision-language models (VLMs) such as CLIP are trained via contrastive learning between text and image pairs, resulting in aligned image and text embeddings that are useful for man…