A*-Thought-V2: Efficient Latent Reasoning via Geometric Dynamics of LLM Paper • 2609.07821 • Published 11 days ago • 18
RealSWE: A Compositional Evaluation of Coding Agents under Realistic User Requests Paper • 2608.27831 • Published 18 days ago • 34
Terminal-Universe: Turning Agent Trajectories into Scalable Terminal Environments Paper • 2609.04148 • Published 15 days ago • 296
Knowing When Not to Reuse: Conditional Experience Transfer in Autonomous LLM Post-Training Paper • 2608.26730 • Published 22 days ago • 157