8 papers · 1 filter
The Mirrored Influence Hypothesis: Efficient Data Influence Estimation by Harnessing Forward Passes
Myeongseob Ko, Feiyang Kang, Weiyan Shi +3
Large-scale black-box models have become ubiquitous across numerous applications. Understanding the influence of individual training data sources on predictions made by these model…
Adversarial Déjà Vu: Jailbreak Dictionary Learning for Stronger Generalization to Unseen Attacks
Mahavir Dabas, Tran Huynh, Nikhil Reddy Billa +8
Large language models remain vulnerable to jailbreak attacks that bypass safety guardrails to elicit harmful outputs. Defending against novel jailbreaks represents a critical chall…
Probing Knowledge Holes in Unlearned LLMs
Myeongseob Ko, Hoang Anh Just, Charles Fleming +2
Machine unlearning has emerged as a prevalent technical solution for selectively removing unwanted knowledge absorbed during pre-training, without requiring full retraining. While…
DiPT: Enhancing LLM reasoning through diversified perspective-taking
Hoang Anh Just, Mahavir Dabas, Lifu Huang +2
Existing work on improving language model reasoning typically explores a single solution path, which can be prone to errors. Inspired by perspective-taking in social studies, this…
Data-Centric Human Preference with Rationales for Direct Preference Alignment
Hoang Anh Just, Ming Jin, Anit Sahu +2
Aligning language models with human preferences through reinforcement learning from human feedback is crucial for their safe and effective deployment. The human preference is typic…
Just Enough Shifts: Mitigating Over-Refusal in Aligned Language Models with Targeted Representation Fine-Tuning
Mahavir Dabas, Si Chen, Charles Fleming +2
Safety alignment is crucial for large language models (LLMs) to resist malicious instructions but often results in over-refusals, where benign prompts are unnecessarily rejected, i…