EvoHarnessRL: Learning Self-Evolving Runtime Harness for Long-Horizon LLM Agents

3 points | by matt_d 11 hours ago