Min Zhang

This paper introduces a novel backdoor attack (ACB) against Reinforcement Learning with Verifiable Rewards (RLVR), demonstrating that poisoning the training data can implant a backdoor that significantly degrades the LLM's safety performance.

The paper introduces SADBench, a systematic benchmark designed to evaluate both the effectiveness of steganographic attacks injecting harmful content and the robustness of steganalysis defenses against these threats.

The paper introduces MARS, a novel meta-adversarial framework that significantly improves black-box adversarial attacks against state-of-the-art Singing Voice Deepfake Detection (SVDD) systems by escaping the Linearity Trap.

The paper proposes Reasoning-Conditioned Direct Preference Optimization (RC-DPO) to effectively mitigate hallucinations in multimodal large reasoning models by explicitly conditioning the preference optimization on the Chain-of-Thought (CoT) process.

The paper introduces Loong, a novel human-like agent that significantly improves long document translation by adaptively selecting and utilizing optimal historical context using a specialized memory module and reinforcement learning.

The paper introduces PersTurnBench, a novel benchmark and evaluator for assessing personalized user conversation satisfaction at specific turns, addressing the limitation of generic response quality metrics.

The paper proposes RIEQE, a two-stage training framework that synergistically co-evolves implicit and explicit reasoning capabilities in Large Reasoning Models (LRMs) to significantly improve fine-grained translation quality estimation.

The paper introduces LongJudgeBench, a new benchmark designed to evaluate the reliability of LLM judges specifically for complex, long-form output evaluation, revealing significant instability gaps in current LLM judging methods.

The paper introduces a novel Clean-Referenced Feature-Vocoder Attack, a black-box adversarial attack that perturbs high-level SSL feature representations instead of raw audio waveforms, achieving superior transferability and robustness against modern ASR defenses.