Skip to main content
Search

# SAKIKO

1 related articles

Subscribe to updates
  1. arXiv

    SAKIKO audits tool-call interventions across seven LLMs: behavioral movement is not repair, and only Qwen3-8B earns a formal ADMIT

    The work presents SAKIKO, an auditing framework that decomposes internal activation interventions on pre-execution tool decisions into directional error discovery, router-conditioned intervention, destination-resolved verification, and prospectively frozen statistical licensing; across seven LLMs on When2Call and MetaTool, channel-keyed interventions induce direction-specific net gains in five models, yet destination auditing shows behavioral movement is not repair—an intervention with +55 net gain corrupts over half of the baseline-correct decisions it touches, Qwen3-4B and Gemma-2-9B are formally declined due to finite-sample uncertainty, and only Qwen3-8B passes all ten conditions for an ADMIT.

You’re all caught up