Vijay Janapa Reddi 72eb537ded feat(vault): make teaching_power gate Bloom-graded (recall is valid at L1) ...

StaffML is an assessment instrument, not a textbook: recall is a first-class
skill and L1/L2 questions are warm-up/screening (paper sec:levels). A flat
'must reason' bar wrongly rejects valid recall items. Grade the floor by Bloom
level instead — L1/L2 pass at tp>=1, L3 at tp>=2, L4/L5/L6+ at tp>=3 — so the
gate fails only a level/depth MISMATCH (recall masquerading as L5), not recall
doing its job. Re-applied to the backfill: 2,700 below-floor (28%, concentrated
at L4-L6+), down from a misleading flat 43%.

2026-05-29 09:39:01 -04:00

..

_batching.py

…

_judges.py

…

apply_corrections.py

…

apply_format_skip_level.py

…

apply_math_skip_level.py

…

apply_proposed_chains.py

…

audit_chains_with_gemini.py

…

audit_corpus_batched.py

…

audit_math.py

…

backfill_provenance.py

…

build_chains_with_gemini.py

…

calibrate_chain_embeddings.py

…

check_chain_decay.py

…

check_registry_append_only.py

…

check_reviewer_identity.py

…

check_schema_sync.py

…

cross_encoder_rerank_experiment.py

…

d1-schema.sql

…

diagnose_chain_coverage.py

…

emit_d1_schema.py

…

exemplar_coverage_audit.py

…

fix_missing_metadata.py

…

generate_question_for_gap.py

…

mass_apply_corrections.py

…

merge_audit_runs.py

…

merge_chain_passes.py

…

normalize_chain_positions.py

…

pre_commit_corpus_guard.py

…

promote_drafts.py

…

regenerate_format_markers.py

…

ship_d1.py

…

suggest_exemplars.py

…

summarize_audit.py

…

summarize_proposed_chains.py

…

validate_drafts.py

…

verify_math_corrections.py

…