Author

Yang Li

Observed name: Li, Yang

Institutions: Rutgers, The State University of New Jersey, Hunan University, Salesforce (United States), University of Hong Kong

Observed institutions: Rutgers, The State University of New Jersey, Hunan University, Salesforce (United States), University of Hong Kong

OpenAlex ID: A5153432674

Broken Symmetry in BF16 Attention: Why FlashAttention Gradients Blow Up Late in Training

2026 · Chen, Junlin, Dong, Daize, et al. · arXiv (Cornell University) · 1

ASCT: Attentive Search over Counterfactual Trees for Credit Assignment in Agentic Reinforcement Learning

2026 · Li, Yang, Yang, Jinhan, et al. · arXiv (Cornell University)

Next Thoughts Are Distributions: Generative Autoregressive Reasoning in the Latent Space

2026 · Li, Yang, Wang, Yi, et al. · arXiv (Cornell University)

Seeing and Solving Are Not Enough for Vision-Language Models

2026 · Wang, Ziheng, Xie, Mingxuan, et al. · arXiv (Cornell University)

Understanding the Synergy between SFT, RLVR, and OPD in LLM Post-Training

2026 · Acikgoz, Emre Can, Li, Yang, et al. · arXiv (Cornell University)

Works count: 5

Citation count: 1

H-index: 1

I10-index: