8 papers
The Paradox of Outcome Optimization: A Causal Information-Theoretic Bound on Reasoning Shortcuts in LLMs
Zihan Chen, Yiming Zhang, Wenxiang Geng +2
Large Language Models (LLMs) aligned via outcome-based Reinforcement Learning (RL) frequently exhibit a critical failure mode: they achieve high performance on in-distribution benc…
Rethinking RL Evaluation: Can Benchmarks Truly Reveal Failures of RL Methods?
Zihan Chen, Yiming Zhang, Hengguang Zhou +3
Current benchmarks are inadequate for evaluating progress in reinforcement learning (RL) for large language models (LLMs).Despite recent benchmark gains reported for RL, we find th…
DINORANKCLIP: DINOv3 Distillation and Injection for Vision-Language Pretraining with High-Order Ranking Consistency
Shuyang Jiang, Nan Yu, Yiming Zhang +2
Contrastive language-image pretraining (CLIP) suffers from two structural weaknesses: the symmetric InfoNCE loss discards the relative ordering among unmatched in-batch pairs, and…
CircuitProbe: Tracing Visual Temporal Evidence Flow in Video Language Models
Yiming Zhang, Zhuokai Zhao, Chengzhang Yu +8
Autoregressive large vision--language models (LVLMs) interface video and language by projecting video features into the LLM's embedding space as continuous visual token embeddings.…
FRAME: Feedback-Refined Agent Methodology for Enhancing Medical Research Insights
Chengzhang Yu, Yiming Zhang, Zhixin Liu +3
The automation of scientific research through large language models (LLMs) presents significant opportunities but faces critical challenges in knowledge synthesis and quality assur…
OrthAlign: Orthogonal Subspace Decomposition for Non-Interfering Multi-Objective Alignment
Liang Lin, Zhihao Xu, Junhao Dong +10
Large language model (LLM) alignment faces a critical dilemma when addressing multiple human preferences: improvements in one dimension frequently come at the expense of others, cr…