4 papers
Beyond Imitation: Auditing the Recoverability of Reasoning in Distilled Models
Ruitong Li, Binjie Guo, Aisheng Mo +4
A correct teacher solution becomes useful supervision when the receiving student can continue its reasoning. We measure this compatibility with prefix recovery: after revealing 25%…
The Order Is the Guarantee: Verifier-Budgeted Code Deletion with Static-First Learned Proposals
Ruitong Li, Binjie Guo, Aisheng Mo +4
Frontier coding models now match or exceed strong human reference points on programming benchmarks, yet benchmark success does not imply maintainable software. Prompt-driven "vibe…
Caliber: Cross-Architecture Extraction-Cost Control for Score-Returning APIs
Chi Wang, Hanwen Wang, Yu Xia +2
We present Caliber, an output-perturbation defense against model extraction that formulates noise selection as a calibration problem: how much the defense degrades the supervision…
When Search Teaches Style: Causal Internalization of Tactical Priors in AlphaZero
Ruitong Li, Binjie Guo, Aisheng Mo +4
AlphaZero is normally evaluated as one agent: a policy-value network fused with Monte Carlo tree search. That fusion hides a causal question. When self-play search is given a usefu…