21 papers
S-SPPO: Semantic-Calibrated Self-Play Preference Optimization
Xiwen Chen, Wenhui Zhu, Jingjing Wang +13
Aligning Large Language Models (LLMs) with human preferences is often formulated via Direct Preference Optimization (DPO). However, the standard Bradley-Terry instantiation of DPO…
DRA-GRPO: Your GRPO Needs to Know Diverse Reasoning Paths for Mathematical Reasoning
Xiwen Chen, Wenhui Zhu, Peijie Qiu +7
Post-training LLMs with Reinforcement Learning, specifically Group Relative Policy Optimization (GRPO), has emerged as a paradigm for enhancing mathematical reasoning. However, sta…
Mags-RL: Wearing Multimodal LLMs a Magnifying Glass via Agentic Reinforcement Learning For Complex Scene Reasoning
Xuanzhao Dong, Wenhui Zhu, Peijie Qiu +11
Despite their popularity and success, Multimodal Large Language Models (MLLMs) often struggle to interpret images accurately, which limits their reasoning capability in complex sce…
SODA: Semi On-Policy Black-Box Distillation for Large Language Models
Xiwen Chen, Jingjing Wang, Wenhui Zhu +7
Black-box knowledge distillation for large language models presents a strict trade-off. Simple off-policy methods (e.g., sequence-level knowledge distillation) struggle to correct…
Your Agent is More Brittle Than You Think: Uncovering Indirect Injection Vulnerabilities in Agentic LLMs
Wenhui Zhu, Xuanzhao Dong, Xiwen Chen +7
The rapid deployment of open-source frameworks has significantly advanced the development of modern multi-agent systems. However, expanded action spaces, including uncontrolled pri…
OTPrune: Distribution-Aligned Visual Token Pruning via Optimal Transport
Xiwen Chen, Wenhui Zhu, Gen Li +9
Multi-modal large language models (MLLMs) achieve strong visual-language reasoning but suffer from high inference cost due to redundant visual tokens. Recent work explores visual t…