This is a heavily interactive web application, and JavaScript is required. Simple HTML interfaces are possible, but that is not what this is.
Post
Sumit
reachsumit.com
did:plc:iakjx7d3ypfa5ycegrylvt2m
Meta-Reinforcement Learning with Self-Reflection for Agentic Search
Introduces a meta-RL framework where search agents learn to improve exploration across episodes via self-reflection.
š arxiv.org/abs/2603.11327
šØš½āš» https://github.com/tengxiao1/MR-Search
https://arxiv.org/abs/2603.11327
2026-03-13T02:35:17.955Z