anhtld commited on
Commit
e323159
·
verified ·
1 Parent(s): 23ebbc0

auto-sync 2026-07-04T04:28:21Z workspace

Browse files
Files changed (1) hide show
  1. workspace/README.md +9 -4
workspace/README.md CHANGED
@@ -263,12 +263,14 @@ CTT training/proxy/rollout:
263
  Use `--no-markdown-report` for README-only runs.
264
  - `scripts/eval_ctt_generated_rollout.py`: measured rollout harness that
265
  restores states, decodes generated tangents, executes candidates, and writes
266
- measured candidate rows.
267
- - `scripts/eval_ctt_rollout.py`: measured-output wrapper.
 
268
  - `scripts/build_ctt_proxy_comparison.py`: proxy comparison/gate table with
269
  by-task/by-seed JSON outputs. Use `--no-markdown-report` for README-only
270
  runs.
271
  - `scripts/build_ctt_rollout_comparison.py`: measured rollout aggregation.
 
272
  - `scripts/build_ctt_outcome_gate.py`: measured outcome acceptance gate that
273
  combines rollout support and train-clean selector artifacts.
274
  - `scripts/summarize_ctt_runs.py`: global CSV/Markdown summary. Markdown output
@@ -280,7 +282,8 @@ Dominance and utility:
280
  - `scripts/calibrate_dominance.py`: conformal-style dominance calibration rule.
281
  - `scripts/eval_dominance_selector.py`: LCB dominance fallback evaluation.
282
  Reports selected success, coverage, fallback, unsafe execution, PCCE,
283
- selector regret, and support/selector gaps.
 
284
  - `scripts/eval_learned_dominance_selector.py`: ridge learned dominance
285
  selector with basic/context/tangent/source/chart-compat and score-shape
286
  features. Use `--no-markdown-report` for README-only runs.
@@ -289,7 +292,9 @@ Dominance and utility:
289
 
290
  Metric and paper artifacts:
291
 
292
- - `scripts/eval_metrics.py`: canonical measured/proxy metric evaluator.
 
 
293
  - `scripts/audit_ctt_paper_artifacts.py`: claim-to-artifact audit for the CTT
294
  paper. It scans forbidden wording, paper table inputs, implementation paths,
295
  and run artifact contracts, then writes JSON/TeX audit outputs without
 
263
  Use `--no-markdown-report` for README-only runs.
264
  - `scripts/eval_ctt_generated_rollout.py`: measured rollout harness that
265
  restores states, decodes generated tangents, executes candidates, and writes
266
+ measured candidate rows. Use `--no-markdown-report` for README-only runs.
267
+ - `scripts/eval_ctt_rollout.py`: measured-output wrapper. Use
268
+ `--no-markdown-report` for README-only runs.
269
  - `scripts/build_ctt_proxy_comparison.py`: proxy comparison/gate table with
270
  by-task/by-seed JSON outputs. Use `--no-markdown-report` for README-only
271
  runs.
272
  - `scripts/build_ctt_rollout_comparison.py`: measured rollout aggregation.
273
+ Use `--no-markdown-report` for README-only runs.
274
  - `scripts/build_ctt_outcome_gate.py`: measured outcome acceptance gate that
275
  combines rollout support and train-clean selector artifacts.
276
  - `scripts/summarize_ctt_runs.py`: global CSV/Markdown summary. Markdown output
 
282
  - `scripts/calibrate_dominance.py`: conformal-style dominance calibration rule.
283
  - `scripts/eval_dominance_selector.py`: LCB dominance fallback evaluation.
284
  Reports selected success, coverage, fallback, unsafe execution, PCCE,
285
+ selector regret, and support/selector gaps. Use `--no-markdown-report` for
286
+ README-only runs.
287
  - `scripts/eval_learned_dominance_selector.py`: ridge learned dominance
288
  selector with basic/context/tangent/source/chart-compat and score-shape
289
  features. Use `--no-markdown-report` for README-only runs.
 
292
 
293
  Metric and paper artifacts:
294
 
295
+ - `scripts/eval_metrics.py`: canonical measured/proxy metric evaluator. It
296
+ writes JSON, TeX, config, command, hash, and log artifacts; use
297
+ `--no-markdown-report` for README-only runs.
298
  - `scripts/audit_ctt_paper_artifacts.py`: claim-to-artifact audit for the CTT
299
  paper. It scans forbidden wording, paper table inputs, implementation paths,
300
  and run artifact contracts, then writes JSON/TeX audit outputs without