Profiles2024-2026
Chao Yang
Shanghai AI Laboratory
15Papers
5Last author
23Unique co-authors
0First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers15
- MCP-SafetyBench: A Benchmark for Safety Evaluation of Large Language Models with Real-World MCP ServersAuthor 5 of 5ICLR 2026 · 34 citations
- Native Reasoning Models: Training Language Models to Reason on Unverifiable DataAuthor 5 of 5ICLR 2026 · 3 citations
- Reflector: Internalizing Step-wise Reflection against Indirect JailbreaksAuthor 6 of 6ICML 2026
- SHADOW: Dynamic-Aware Credit Assignment Against Long-Horizon TasksAuthor 3 of 3AAAI 2026
- TreeEval: Benchmark-Free Evaluation of Large Language Models through Tree PlanningAuthor 3 of 3AAAI 2025 · 17 citations
- Adversarial Preference Learning for Robust LLM AlignmentAuthor 8 of 16ACL Findings 2025 · 4 citations
- Emergent Response Planning in LLMsAuthor 4 of 5ICML 2025
- Weak-to-Strong Search: Align Large Language Models via Searching over Small Language ModelsAuthor 5 of 6NeurIPS 2024 · 50 citations
- Critic-Guided Decision Transformer for Offline Reinforcement LearningAuthor 2 of 5AAAI 2024 · 35 citations
- Beyond One-Preference-Fits-All Alignment: Multi-Objective Direct Preference OptimizationAuthor 5 of 7ACL Findings 2024 · 21 citations
- LLaMA-Excitor: General Instruction Tuning via Indirect Feature InteractionAuthor 2 of 5CVPR 2024 · 5 citations
- SEER: Facilitating Structured Reasoning and Explanation via Reinforcement LearningAuthor 3 of 6ACL 2024 · 5 citations
- Language-aware Visual Semantic Distillation for Video Question AnsweringAuthor 2 of 5CVPR 2024 · 2 citations
- Inference-Time Language Model Alignment via Integrated Value GuidanceAuthor 4 of 5EMNLP Findings 2024 · 1 citation
- Emulated Disalignment: Safety Alignment for Large Language Models May Backfire!Author 5 of 7ACL 2024
