feat(eval): Phase 13.5 classification framework freeze + true end-to-end evaluator - #12

Merged

xianyu9n merged 6 commits into

phase13-5-e2e-freeze

Aug 19, 2026

Commits on Aug 19, 2026

feat(eval): true end-to-end choice-protocol evaluator + unit tests; classification freeze doc (phase 13.5)
曾立宏
committed
da8252d
docs(freeze): add real per-stage I/O examples and true-E2E success/fail traces (pers_info)
曾立宏
committed
7ed073d
docs(freeze): label the field metadata as the data-to-classify + raw->prompt visibility mapping
曾立宏
committed
2f653f2
docs(freeze): narrow freeze scope to classification pipeline; data_level/grading declared not-frozen (pending clarification)
曾立宏
committed
6b8714c
docs(freeze): PR review - unfreeze synthetic candidate fixture & provisional reward coefficient; grading additive-extension; RL-init wording (no specific checkpoint frozen)
曾立宏
committed
6a11561
docs(freeze): check_stage2_choices stays frozen classification-only strict parser; grading resolved by a new joint parser reusing answer-choice decode
曾立宏
committed
c8e38c9