REVEAL: Reverse Engineering Agentic Latents from Frontier-Model Traces
Frontier LLM agents perform complex interactive tasks, but their inference costs motivate transferring their experience to smaller open-source agents. Distillation and latent communication typically require target-model training or access to source-model internals. We study how textual frontier-model trajectories can p...
Yue-Tian Du, Meng Wang, Yu-Cheng Wang et al.
· 0 citations