AutoFyn Technical Report: Non-Parametric Expert Iteration for Long-Horizon Agents
Source
Adib Hasan, Daniel Schaffield, Akashnil Dutta, Tarik Adnan Moon
Author
Adib Hasan, Daniel Schaffield, Akashnil Dutta, Tarik Adnan Moon
Date
Terms in this piece · Glossary
agent harness — The scaffolding around a model that turns it into a working agent — the loop, the tools it can call, and the rules for when to stop.
Why it matters
Instead of retraining weights, this agent harnessThe scaffolding around a model that turns it into a working agent — the loop, the tools it can call, and the rules for when to stop.Full definition → distills verified rewards each round into persistent memory files and repo state a fresh model session reads back, tested on IMO 2026, data science, and cybersecurity tasks.
articleAutoWorldModel-Bench: A State-Centric Benchmark for Automated World-Model ResearchMarjan Moodi, Xuankang Zhu, Fernando De Mesentier Silva, Harold Chaput, Mohammad Reza Taesiri
articleSelf-Evolving Embodied Agents via Skill-Harness EvolutionPeidong Wang, Zhiming Ma, Ying Chang, Xufang Luo, Xiaocui Yang, Shi Feng, Yuqing Yang, Dongsheng Li
articleFresh Memory, Stale Plans: Dependency-Scoped Validation for Distributed LLM-Agent MemoryEvan Chen, Shiqiang Wang, Christopher G. Brinton