AI RESEARCH
Bounded Behavioral Indistinguishability for Black-Box LLM Distillation
arXiv CS.LG
•
ArXi:2605.30448v1 Announce Type: new Black-box LLM distillation is usually evaluated as an output-matching problem: a student is considered successful when its responses are semantically similar to, or task-consistent with, those of a teacher. However, output similarity does not imply that the student is behaviorally indistinguishable from the model it imitates. We We instantiate this notion on Qwen and Llama teacher-student pairs using a controlled $5,000$-prompt behavioral probe suite.