…We train a lightweight sampling controller with reinforcement learning (RL) to jointly balance answer correctness, latency, and computation cost. At each round, the controller decides to stop sampling or to acquire additional…
…Visualization is most useful during fault localization and when the agent autonomously controls exploration depth . These findings point to a practical hybrid text-and-vision design for next-generation coding agents. View…
…improving agent safety comes at the cost of degraded task performance . Such sparse and single-objective rewards severely limit real-world usability. To bridge this gap, we propose FATE, an on-policy…
…a Long-Horizon Memory Environment for LLM Agents (2026) MementoGUI: Learning Agentic Multimodal Memory Control for Long-Horizon GUI Agents (2026) When Stored Evidence Stops Being Usable: Scale-Conditioned Evaluation of Agent…
…This closed-loop system transforms real-world manipulation learning into a controllable optimization procedure, minimizing human effort while allowing fair ablations across training recipe and agent variants. Powered by ENPIRE, frontier coding…
…By connecting visual observations with human actions, camera motion, game states, and events at scale, EgoCS-400K serves as a practical bridge between passive web videos, controllable game simulation, and costly real…
…A strong coding agent today can probably generate data by inspecting a schema, but PIPE-Cypher makes this scalable, cost-effective, and repeatable when the schema inevitably changes. By constraining this as…
…Work focuses on improving the efficiency and robustness of distributed black box optimization in multi-agent systems. Potential applications include cooperative sensing, resource allocation, and distributed control, which may contribute to more…
…controller to optimize routing and reduce latency. Generated by Qwen/Qwen2.5-Coder-32B-Instruct Disaggregated inference architectures physically separate prefill and decode phase s onto distinct GPU pools , creating competing "agents…
…Only a file-based agent paired with Claude Opus 4.7 as its internal LLM partially closes the gap, but at ~70x the baseline cost, indicating closure currently depends on configurations that…