Tianfu Fu

2 accepted papers

2026

AutoCode: LLMs as Problem Setters for Competitive Programming

Shang Zhou, Zihan Zheng, Kaiyuan Liu, Zeyu Shen, Zerui Cheng, Zexing Chen +15

ICLR 2026poster

Writing competitive programming problems is exacting. Authors must: set constraints, input distributions, and edge cases that rule out shortcuts; target specific algorithms (e.g., max-flow, dynamic programming, data structures); and calibrate complexity beyond the reach of most competitors. We argue…

Cited by 0 Source ↗Scholar ↗

2026

TriAttention: Efficient Long Reasoning with Trigonometric KV Compression

Weian Mao, Xi Lin, Wei Huang, Yuxin Xie, Tianfu Fu, Bohan Zhuang +2

ICML 2026poster

Extended reasoning in large language models (LLMs) requires long and accurate decoding and creates severe KV cache memory bottlenecks. Leading KV cache compression methods estimate KV importance using attention scores from recent post-RoPE queries. However, queries rotate with position during RoPE, …

Cited by 0 Source ↗Scholar ↗