41 citations · 72 across the 28 of their papers we have counts for
15 papers · 1 filter
Capturing the Temporal Dependence of Training Data Influence
Jiachen T. Wang, Dawn Song, James Zou +2
Traditional data influence estimation methods, like influence function, assume that learning algorithms are permutation-invariant with respect to training data. However, modern tra…
On Evaluating the Durability of Safeguards for Open-Weight LLMs
Xiangyu Qi, Boyi Wei, Nicholas Carlini +7
Stakeholders -- from model developers to policymakers -- seek to minimize the dual-use risks of large language models (LLMs). An open challenge to this goal is whether technical sa…
Instructional Segment Embedding: Improving LLM Safety with Instruction Hierarchy
Tong Wu, Shujian Zhang, Kaiqiang Song +7
Large Language Models (LLMs) are susceptible to security and safety threats, such as prompt injection, prompt extraction, and harmful requests. One major cause of these vulnerabili…
Lottery Ticket Adaptation: Mitigating Destructive Interference in LLMs
Ashwinee Panda, Berivan Isik, Xiangyu Qi +3
Existing methods for adapting large language models (LLMs) to new tasks are not suited to multi-task adaptation because they modify all the model weights -- causing destructive int…
Safety Alignment Should Be Made More Than Just a Few Tokens Deep
Xiangyu Qi, Ashwinee Panda, Kaifeng Lyu +5
The safety alignment of current Large Language Models (LLMs) is vulnerable. Relatively simple attacks, or even benign fine-tuning, can jailbreak aligned models. We argue that many…
SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal
Tinghao Xie, Xiangyu Qi, Yi Zeng +13
Evaluating aligned large language models' (LLMs) ability to recognize and reject unsafe user requests is crucial for safe, policy-compliant deployments. Existing evaluation efforts…
Data Shapley in One Training Run
Jiachen T. Wang, Prateek Mittal, Dawn Song +1
Data Shapley provides a principled framework for attributing data's contribution within machine learning contexts. However, existing approaches require re-training models on differ…
AI Risk Management Should Incorporate Both Safety and Security
Xiangyu Qi, Yangsibo Huang, Yi Zeng +22
The exposure of security vulnerabilities in safety-aligned language models, e.g., susceptibility to adversarial attacks, has shed light on the intricate interplay between AI safety…
Beyond Performance: Quantifying and Mitigating Label Bias in LLMs
Yuval Reif, Roy Schwartz
Large language models (LLMs) have shown remarkable adaptability to diverse tasks, by leveraging context prompts containing instructions, or minimal input-output examples. However,…
Certifiably Robust RAG against Retrieval Corruption
Chong Xiang, Tong Wu, Zexuan Zhong +3
Retrieval-augmented generation (RAG) is susceptible to retrieval corruption attacks, where malicious passages injected into retrieval results can lead to inaccurate model responses…
Position: Towards Resilience Against Adversarial Examples
Sihui Dai, Chong Xiang, Tong Wu +1
Current research on defending against adversarial examples focuses primarily on achieving robustness against a single attack type such as or -bounded attack…
Teach LLMs to Phish: Stealing Private Information from Language Models
Ashwinee Panda, Christopher A. Choquette-Choo, Zhengming Zhang +2
When large language models are trained on private data, it can be a significant privacy risk for them to memorize and regurgitate sensitive information. In this work, we propose a…
Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modifications
Boyi Wei, Kaixuan Huang, Yangsibo Huang +6
Large language models (LLMs) show inherent brittleness in their safety mechanisms, as evidenced by their susceptibility to jailbreaking and even non-malicious fine-tuning. This stu…
Efficient Data Shapley for Weighted Nearest Neighbor Algorithms
Jiachen T. Wang, Prateek Mittal, Ruoxi Jia
This work aims to address an open problem in data valuation literature concerning the efficient computation of Data Shapley for weighted nearest neighbor algorithm (WKNN-Shaple…
Private Fine-tuning of Large Language Models with Zeroth-order Optimization
Xinyu Tang, Ashwinee Panda, Milad Nasr +2
Differentially private stochastic gradient descent (DP-SGD) allows models to be trained in a privacy-preserving manner, but has proven difficult to scale to the era of foundation m…