1 paper · 2 filters
Caiqi Zhang, Rujun Han, Zifeng Wang +4
As LLM agents undertake increasingly complex, long-horizon tasks, verifying their outputs becomes increasingly challenging. We study how verification capability can be strengthened…