인터페이스는 현지화되어 있으며, Skill 소개와 설명 문서는 선택한 언어로 AI가 자동 번역합니다. 코드·명령어·경로·전문 용어는 원문을 유지하며 원작자 문서가 우선합니다.
Skills 둘러보기
생산성 및 협업 · SKILL.md

tsnpe_reduced_recovery_harness

tsnpe_reduced_recovery_harness

Run a bounded mechanism-faithful TSNPE proxy recovery experiment.

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill 문서

GitHub

name: tsnpe_reduced_recovery_harness description: Run a bounded mechanism-faithful TSNPE proxy recovery experiment.

Tsnpe Reduced Recovery Harness

Use this skill when full TSNPE benchmark recovery is blocked by heavy package, simulator, or data requirements and soft mode permits a declared proxy. The harness must call the generated HPR, training, and diagnostic scripts, produce executable evidence, and label the result as reduced/proxy.

Inputs: attempt directory, module plan, runtime handoff, and generated skill root. Outputs: recovery result and logs under the attempt recovery directory. Validation: run the harness followed by the Distiller recovery validator. For reduced Gaussian recovery, accept the proxy only when loss decreases, ground-truth support inclusion is at least 0.5, absolute mean error is finite and below 1.0, and all mechanism checks remain true.

저장한 Skills