arXiv Machine Learning
Fork-and-Flush: Escaping Idea Basins in Autoresearch Agents
Read the original on arXiv Machine Learning →The Flow has not summarised this story yet — read it at arXiv Machine Learning.
Related stories
arXiv Computation and Language
Oct 12
Scaling Agentic Data for Long-Horizon Terminal Intelligence
The paper introduces TerminalHorizon, an automated engine that synthesizes long‑horizon trajectories for terminal agents by reconstructing environments from real‑world professional work and progressively expanding tasks to require further exploration and improvement. Using this engine, the authors create 1.5K task environments and fine‑tune Qwen3.5‑27B, producing TerminalHorizon‑27B, which shows significant performance gains across 12 benchmarks in coding, agentic production tasks, and reasoning. Case studies demonstrate the model’s ability to experiment beyond initial solutions, refine methods based on feedback, and preserve effective improvements in final deliverables.
arXiv AI
Jul 7
Automating the Design of Embodied Agent Architectures
arXiv:2606. 30111v2 Announce Type: replace-cross Abstract: Embodied agents are typically built as hand-designed compositions of perception, memory, planning, and action modules.
arXiv AI
Oct 7
Stateless Language Agents: Scaling Long-Horizon Automated Research
arXiv:2610.07625v1 Announce Type: cross Abstract: Automated research systems increasingly run LLM agents over long horizons, but more inference does not by itself produce more progress: agents replay...
arXiv AI
Jun 30
Automating the Design of Embodied AgentArchitectures
arXiv:2606. 30111v1 Announce Type: cross Abstract: Embodied agents are typically built as hand-designed compositions of perception, memory, planning, and action modules.
arXiv Computer Vision
Aug 28
OS-Marathon: Benchmarking Computer-Use Agents on Vast-Horizon, Repetitive Tasks
OS-Marathon is a new benchmark that tests computer‑use agents on vast‑horizon, repetitive tasks, covering 100 tasks across five scenarios and ten domains. The study shows that current state‑of‑the‑art agents perform poorly on these tasks, and that simply decomposing workflows into subtasks does not solve the problem. Introducing a cost‑friendly personalization method called GraphDemo, which adapts agents from a single human demonstration, improves performance, highlighting the value of human guidance for these challenging tasks.
arXiv AI
Jul 28
A Vocabulary for Multi-Agent Automated Research Systems
arXiv:2607. 22682v1 Announce Type: new Abstract: We introduce a vocabulary for automated research systems built from one or more agents to make their design choices easier to describe and compare.