Sitemap
A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.
Pages
Posts
Humans Still Beat AI in the Long Horizon: Revisiting Test-Time Scaling in the Agent Era
Published:
Agents can spend test-time compute by trying, observing, and revising. We derive an Elo reference for repeated sampling, then show that in a 2022 two-week coding marathon, current agents plateau within 24 hours while top humans keep improving.
Your Next Long-Context Recipe: Open-Ended Problems
Published:
We integrate FrontierCS into Harbor and release a preview long-horizon agent leaderboard on 178 open-ended algorithmic tasks. Kimi K2.6 and Claude Code Opus 4.7 show similar headline capability, but very different failure modes.
一段故事的句号
Published:
我曾无数次向大海呼唤，
portfolio
Portfolio item number 1
Published:
Short description of portfolio item number 1
Portfolio item number 2
Published:
Short description of portfolio item number 2
publications
LiveCodeBench Pro: How Do Olympiad Medalists Judge LLMs in Competitive Programming?
Published in Neurips 2025, 2025
Authors: Zihan Zheng*, Zerui Cheng*, Zeyu Shen*, Shang Zhou*, Kaiyuan Liu*, et al.
AutoCode: LLMs as Problem Setters for Competitive Programming
Published in ICLR 2026, 2025
Authors: Shang Zhou*, Zihan Zheng*, Kaiyuan Liu*, et al.
FrontierCS: Evolving Challenges for Evolving Intelligence
Published in ICML 2026, 2025
Authors: Qiuyang Mang*, Wenhao Chai*, Zhifei Li*, Huanzhi Mao*, Shang Zhou*, …, Kaiyuan Liu, et al.
Evaluating LLM Agents as Human Simulators in Climate Social Dilemmas
Published in ICLR 2026 MALGAI Workshop (Oral), 2026
Authors: Kaiyuan Liu*, Xiaoxuan Hou*, Carrie Yuan, Natasha Jaques.
FrontierSmith: Synthesizing Open-Ended Coding Problems at Scale
Published in NeurIPS 2026 (Spotlight), 2026
Authors: Runyuan He, Qiuyang Mang, Shang Zhou, Kaiyuan Liu, Hanchen Li, Huanzhi Mao, et al.
OpenDeepThink: Parallel Reasoning via Bradley-Terry Aggregation
Published in Under Review at ICLR 2027, 2026
Authors: Shang Zhou, Wenhao Chai, Kaiyuan Liu, Huanzhi Mao, Qiuyang Mang, Jingbo Shang
Agents’ Last Exam
Published in NeurIPS 2026, 2026
Authors: Yiyou Sun, Xinyang Han, Weichen Zhang, …, Kaiyuan Liu, …, Dawn Song
Harbor Adapters and Harbor-Index: Infrastructure and a Curated Meta-Dataset for Large-Scale Agentic Evaluation
Published in NeurIPS 2026, 2026
Authors: Lin Shi, Haowei Lin, Zixuan Zhu, …, Kaiyuan Liu, et al.
When Agents Slow Down: Understanding LLM Agents’ Test-Time Strategies via Elo-per-token Analysis
Published in arXiv Preprint, 2026
Authors: Kaiyuan Liu, Qiuyang Mang, Bo Peng, Wenhao Chai, Hanchen Li, Shreyas Pimpalgaonkar, Luke Zettlemoyer, Alex Dimakis, Alvin Cheung
AutoCode-RL: Climbing the Difficulty Pyramid with Verifiable Synthetic Problems
Published in Under Review at ICLR 2027, 2026
Authors: Kaiyuan Liu, Zhiming Wang, Zeyu Shen, et al.
EmbodiedSWE: Coding Agents for Long Horizon Dexterous Robotics
Published in arXiv Preprint, 2026
Authors: Zeyu Shen, Haoxiang You, Yilang Liu, …, Kaiyuan Liu, et al.
FrontierCS 2: A Benchmark for Computer Science R&D
Published in arXiv Preprint, 2026
Authors: Qiuyang Mang, Wenhao Chai, Runyuan He, …, Kaiyuan Liu, et al.
talks
Talk 1 on Relevant Topic in Your Field
Published:
This is a description of your talk, which is a markdown files that can be all markdown-ified like any other post. Yay markdown!
Conference Proceeding talk 3 on Relevant Topic in Your Field
Published:
This is a description of your conference proceedings talk, note the different field in type. You can put anything in this field.
teaching
CSE 417
Undergraduate course, University of Washington, Paul Allen School, 2014
This is a description of a teaching experience. You can use markdown like any other post.
Teaching experience 2
Workshop, University 1, Department, 2015
This is a description of a teaching experience. You can use markdown like any other post.