Lune

Profiles2024-2026

Yankai Lin

Renmin University of China

13Papers
7Last author
15Unique co-authors
0First author
0Solo

First and last author come from byline position, which some venues assign alphabetically.

Recent papers13

  1. LaSeR: Reinforcement Learning with Last-Token Self-RewardingAuthor 7 of 7ICLR 2026 · 12 citations
  2. CURE: Critique-Driven Unified Reinforcement Learning for Test-Time Self-ImprovementAuthor 6 of 6ACL 2026
  3. Less Noise, More Voice: Reinforcement Learning for Reasoning via Instruction PurificationAuthor 3 of 3ACL Findings 2026
  4. Scaling Spatial Reasoning in MLLMs through Programmatic Data SynthesisAuthor 12 of 13CVPR Findings 2026
  5. Towards Preference Following in Tool Calling Language AgentsAuthor 5 of 5ACL Findings 2026
  6. Towards Thinking-Optimal Scaling of Test-Time Compute for LLM ReasoningAuthor 3 of 4NeurIPS 2025 · 141 citations
  7. Learning to Focus: Causal Attention Distillation via Gradient-Guided Token PruningAuthor 6 of 6NeurIPS 2025 · 14 citations
  8. Beyond the Surface: Measuring Self-Preference in LLM JudgmentsAuthor 5 of 5EMNLP 2025
  9. Rational Decision-Making Agent with Learning Internal Utility JudgmentAuthor 6 of 8ICLR 2025
  10. Super(ficial)-alignment: Strong Models May Deceive Weak Models in Weak-to-Strong GeneralizationAuthor 7 of 8ICLR 2025
  11. Towards Codable Watermarking for Injecting Multi-Bits Information to LLMsAuthor 5 of 8ICLR 2024 · 55 citations
  12. Large Language Model-based Human-Agent Collaboration for Complex Task SolvingAuthor 4 of 7EMNLP Findings 2024 · 24 citations
  13. Towards Tool Use Alignment of Large Language ModelsAuthor 6 of 6EMNLP 2024 · 5 citations
Dusk over the sea between two cliffs drawn in fine vertical lines