1 citations · 2 across the 15 of their papers we have counts for
16 papers · 1 filter
Revisiting Over-refusal in Safety-aligned Large Language Models: A Representation and Optimization Perspective
Chao Chen, Pengyang Shao, Wei Qin +3
Over-refusal is a critical side effect of safety alignment in large language models (LLMs), causing harmless but superficially risky queries to be mistakenly rejected. Despite the…
Safety Kernel: Mechanistic Alignment for Cross-Lingual Defense in LLMs
Chengyu Xie, Fei Shen, Shuyi Miao +5
The safety capabilities of large language models (LLMs) are highly uneven across different languages, often allowing cross-lingual attacks to bypass safety guardrails. Existing ali…
MemForm: Benchmarking Memory Formation Poisoning in LLM Agents
Haowen Pan, Pengyang Shao, Yunshan Ma +4
Long-term memory enables LLM agents to learn from past interactions but exposes memory formation to adversarial influence. Existing studies primarily organize memory poisoning by a…
AlphaPADI: Formulaic Alpha Discovery via Pool-Aware Hierarchical Discrete Diffusion
Yanzheng Jin, Pengyang Shao, Yunshan Ma +5
Formulaic alpha discovery seeks symbolic expressions that predict cross-sectional asset returns. In deployment, multiple formulas are combined into an alpha pool, where each formul…
ACTR: Aligning Thoughts and Responses for Multilingual Safety in Reasoning LLMs
Xianhui Zhang, Jian Yu, Chengyu Xie +6
Ensuring the safety of reasoning large language models (LLMs) across languages is essential for their reliable deployment. However, when exposed to jailbreak attacks in non-high-re…
Hearsay: Can an Auditor Trust the Record a Deployed Agent Harness Writes?
Jiahong Dai, Zhuochen Yang, Pengyang Shao +5
An agent harness, the code that turns a model into an agent, writes its own record of each run, and that record is all a later reader gets when a run is disputed, investigated or a…
Beyond Cross-Lingual Transfer: Benchmarking Propagation Boundaries in Multilingual LLM Unlearning
Pengyang Shao, Chuanpeng Lu, Wei Qin +5
Large Language Model (LLM) unlearning aims to suppress target knowledge while preserving general capabilities. In multilingual settings, unlearning must additionally propagate with…
Who Bridges Safety? Identifying and Targeting Cross-Lingual Shared Safety Pathways
Shuyi Miao, Wangjie Qiu, Pengyang Shao +4
Uncovering the internal mechanisms underlying the safety capabilities of large language models (LLMs) is crucial for developing trustworthy artificial intelligence. Currently, mech…
CoLAS: Multimodal Corroboration of Latent Asset Signals for Financial Trading
Yanzheng Jin, Pengyang Shao, Xiaohao Liu +3
Financial trading relies on extracting reliable signals from heterogeneous market modalities such as price series, breaking news, and investor sentiment. Existing multimodal method…
Sharpness-Aware Poisoning: Enhancing Transferability of Injective Attacks on Recommender Systems
Junsong Xie, Yonghui Yang, Pengyang Shao +1
Recommender Systems~(RS) have been shown to be vulnerable to injective attacks, where attackers inject limited fake user profiles to promote the exposure of target items to real us…
StressWeb: A Diagnostic Benchmark for Web Agent Robustness under Realistic Interaction Variability
Haoyue Bai, Dong Wang, Long Chen +5
Large language model-based web agents have demonstrated strong performance on realistic web interaction tasks. However, existing evaluations are predominantly conducted under relat…
MURE: Hierarchical Multi-Resolution Encoding via Vision-Language Models for Visual Document Retrieval
Fengbin Zhu, Zijing Cai, Yuzhe Wang +5
Visual Document Retrieval (VDR) requires representations that capture both fine-grained visual details and global document structure to ensure retrieval efficacy while maintaining…
CLEAR: Null-Space Projection for Cross-Modal De-Redundancy in Multimodal Recommendation
Hao Zhan, Yihui Wang, Yonghui Yang +6
Multimodal recommendation has emerged as an effective paradigm for enhancing collaborative filtering by incorporating heterogeneous content modalities. Existing multimodal recommen…
CompassDPO: Dynamics-Controlled Direct Preference Optimization for Robust Safety Alignment
Jilong Liu, Yonghui Yang, Pengyang Shao +5
Direct Preference Optimization (DPO) has become a standard framework for safety alignment, but its reliance on pairwise preference updates makes training sensitive to imperfect sup…
BalDRO: A Distributionally Robust Optimization based Framework for Large Language Model Unlearning
Pengyang Shao, Naixin Zhai, Lei Chen +4
As Large Language Models (LLMs) increasingly shape online content, removing targeted information from well-trained LLMs (also known as LLM unlearning) has become critical for web g…
Maximizing Local Entropy Where It Matters: Prefix-Aware Localized LLM Unlearning
Naixin Zhai, Pengyang Shao, Binbin Zheng +4
Machine unlearning aims to forget sensitive knowledge from Large Language Models (LLMs) while maintaining general utility. However, existing approaches typically treat all tokens i…