TRACE: Capability-Targeted Agentic Training
arXiv:2604.05336v1 Announce Type: new Abstract: Large Language Models (LLMs) deployed in agentic environments must exercise multiple capabilities across different task instances, where a capability is …
Hangoo Kang, Tarun Suresh, Jon Saad-Falcon, Azalia Mirhoseini
조회수 71회