core/evals
Shay 0be18ebdb9
feat(reader): add overlaps_event finite-verb reader surface (B3) (#783)
Add a closed finite-verb relational surface to comprehend_relational, narrowly:
declarative <A> overlaps <B>, interrogative Does <A> overlap <B>?, predicate
overlaps_event ONLY (closed finite-verb table, default-off).

Kept SEPARATE from the copula-connective grammar (byte-unchanged): the other
connectives (before/after/during/inside/adjacent/…) still REQUIRE the copula, so
"Monday before Friday." stays a refusal — no connective bypass (adversarial hazard #2).

Fail-closed slot gate: each finite-verb argument slot must be EXACTLY ONE content
token (after article stripping). An enumerated adverb/negation blocklist is unbounded
and leaks — the adversarial audit found "Meeting never overlaps lunch." committed as a
POSITIVE overlaps_event(meeting_never, lunch), plus almost/sometimes/trailing-qualifier/
second-verb compound fabrications in BOTH slots and the query path. The single-token
gate closes the whole COMPOUND-fabrication class (any extra token refuses); the
_FINITE_VERB_MODIFIERS and _CONNECTIVE_TOKENS checks add precise reasons for common
bare-modifier / second-verb slots. Multi-word entities in the finite-verb surface are
deferred (they need a positive content lexicon, which OOV entities preclude). A lone
bare token remains an entity (the reader's universal OOV single-token contract — the
copula and general readers behave identically), not a compound fabrication.

Hazards pinned (adversarial audit hazard #1 + #2): adverb absorption, negation-as-
positive ('never'), interrogative double-verb, trailing qualifier, and the copula-bypass
firewall.

Measurement: migrated ref-009 "Sunrise overlaps dawn." (the documented coverage gap) to
a positive rel-018 — comprehension_relational_predicate 17->18, wrong_total 0, breadth
still 11 (coverage added, not a new domain), baseline digest re-frozen. Refusal floor
rose 9 -> 24 (finite-verb confusers replace the one migrated input). The #775 inference
and #781 transitive lane SHA pins are UNTOUCHED. determine.py is untouched (reader-only):
no answer=False, no FrameVerdict, INV-30 unaffected.

Verified in the worktree: finite-verb + reader + reader-lane + #775 inference lane +
#781 transitive lane + #779 OWA floor + capability baseline/index + INV-30/29/21 + full
smoke — all green. Two read-only adversarial passes: the first found 4 real slot-
fabrication blockers, closed by the single-token gate; a black-box probe of 24
adversarial inputs confirms the compound-fabrication class is fully closed.
2026-06-15 14:56:09 -07:00
..
adversarial_identity
anchor_lens_tour
anti_regression feat(epistemic): Phase 3 state tagging spine (#220) 2026-05-24 11:26:06 -07:00
articulation
articulation_of_status
audio_sensorium feat(adr-0181-p4): audio compiler eval gate lane (sensorium/audio) (#470) 2026-05-29 11:20:31 -07:00
audit_tour
calibration
capability_index feat(reader): add overlaps_event finite-verb reader surface (B3) (#783) 2026-06-15 14:56:09 -07:00
classical_literature_ood
cognition
cold_start_grounding fix(quarantine): drain all 60 quarantined tests — QUARANTINE=∅ (#267) 2026-05-25 11:22:12 -07:00
combined_rate_oracle docs(cmb): lookback review + fix single-agent-attribution hygiene hazard (H3) 2026-06-08 13:27:17 -07:00
compositionality
compound_intent_decomposition
comprehension feat: one-hop sound relational entailment (inverse/symmetric) + capability-index lane (#775) 2026-06-15 11:39:41 -07:00
constraint_oracle feat(verified): P1-C — split bound_slots_digest into separate obligation 2026-06-08 17:17:31 -07:00
contemplation_quality feat(W-025): contemplation quality eval lane (ADR-0159) (#286) 2026-05-25 20:38:52 -07:00
contradiction_detection
conversation
conversational_thread_coherence
cross_domain_transfer
curriculum_loop_closure
deductive_logic test(l10): add independent-gold adversarial logic fixtures 2026-06-05 09:08:23 -07:00
demo_composition fix(W-015): replace _slerp_toward with rotor-geodesic anchor pull (#255) 2026-05-24 20:05:25 -07:00
determination_closure feat(determine): idle deductive consolidation — the loop learns from determined facts (Step D) 2026-06-06 12:28:09 -07:00
determination_estimation feat(determine): calibrated disclosed estimation — the engine earns the right to guess (Step E) 2026-06-06 13:49:07 -07:00
deterministic_fluency
dimensional feat: dimensional-reasoning lane — 3rd diversity-panel domain 2026-06-04 16:38:56 -07:00
discourse_paragraph
domain_contract_validation chore(ci): re-pin drifted lane SHAs + refresh canonical reports (#229) 2026-05-24 14:25:11 -07:00
edge_budget feat(edge): edge-deployment budget gate — deterministic per-turn persistence cost 2026-06-06 10:27:10 -07:00
elementary_mathematics_ood
english_fluency_ood
environment_falsification Add offline witness log importer 2026-06-06 12:37:57 -07:00
event_vision_sensorium Add event vision sensorium lane 2026-06-06 12:37:57 -07:00
fabrication_control chore(ci): re-pin drifted lane SHAs + refresh canonical reports (#229) 2026-05-24 14:25:11 -07:00
flywheel_demo feat(demo): core demo flywheel — public-proof reproduction of the loop 2026-05-27 21:33:54 -07:00
forward_semantic_control
foundational_biology_ood
foundational_physics_ood
frontier_compare
grammatical_coverage
gsm8k_math feat(workbench): land B3.5-b/c/d/e — calibration evidence subject, B4a leeway gate, docs; runner-reproducible practice artifact 2026-06-13 07:36:44 -07:00
gsm8k_parser_dev
hebrew_fluency
identity_divergence
industry_demos
inference_closure
introspection
koine_greek_fluency
l10_always_on feat(identity): split engine identity from build provenance (ADR-0220 PR C) (#774) 2026-06-15 11:38:04 -07:00
l10_continuity test(l10): W2-R arbitrary-interruption recovery harness (ADR-0219) 2026-06-15 02:45:29 -07:00
lab
learning_arc fix(W-025): polish contemplation-quality eval lane follow-ups (#290) 2026-05-26 09:39:18 -07:00
learning_loop
long_context_cost
math_bounded_grammar/v1
math_capability_axes chore(evals): refresh stale committed reports 2026-06-03 01:01:50 -07:00
math_expert_claims/v1 docs(claims): ADR-0200 reconciliation — expert claim to audit-passed truth 2026-06-02 10:06:16 -07:00
math_symbolic_equivalence
math_teaching_corpus/v1
miner_loop_closure
monotonic_learning
multi_agent_composition
multi_sentence_response
multi_step_reasoning
obligation_2_ood_ratio feat(ADR-0114a.2): OOD-ratio auditor — Obligation #2 wired for B3, ratio=1.00 (#193) 2026-05-23 16:25:28 -07:00
obligation_5_perturbation feat(ADR-0114a.5): reasoning-isolation perturbation suite — Obligation #5 wired for B3, PASSING 130/130 preserving, 68/68 breaking (#191) 2026-05-23 16:07:59 -07:00
obligation_6_depth_curve feat(ADR-0114a.6): depth-curve auditor — Obligation #6 wired for B3 (assertion holds, coverage gap named) (#190) 2026-05-23 16:19:58 -07:00
obligation_8_adversarial feat(ADR-0114a.8): adversarial auditor — Obligation #8 wired, PASSING; surfaces 2 known parser-layer gaps (#192) 2026-05-23 16:11:37 -07:00
obligation_10_pack_provenance
orthogonality_tour
prompt_diversity
proofwriter_owa feat(eval): ProofWriter-OWA refusal-floor lane (B1) — independent oracle, measure-only (#779) 2026-06-15 12:28:20 -07:00
propositional_logic feat(comprehend): propositional-logic comprehension (4th domain, flagship oracle) 2026-06-05 23:24:54 -07:00
provenance
public_demo docs(public_demo): add known environment caveat for wall-clock budget flake 2026-06-11 18:12:00 -07:00
rate_oracle test(rate-oracle): port _canonical_outcome non-vacuous validation to R3 (R3-vac) 2026-06-08 07:38:20 -07:00
realizer_guard
refusal_calibration feat(identity): L11 identity continuity — same identity across reboot, not just same bytes 2026-06-05 13:52:57 -07:00
refusal_taxonomy fix(refusal-taxonomy): parse new refusal-reason format; reconcile to reader gains 2026-06-03 01:49:41 -07:00
register_diagnostics
register_tour
relational feat(reader): add overlaps_event finite-verb reader surface (B3) (#783) 2026-06-15 14:56:09 -07:00
relational_inference/v1 feat(reader): add overlaps_event finite-verb reader surface (B3) (#783) 2026-06-15 14:56:09 -07:00
relational_metric feat(oracle): narrow reverse-solve contract — base of one more/fewer_than (PR-7a) 2026-06-07 06:22:36 -07:00
relational_transitive feat(determine): add transitive strict-order relational inference (#781) 2026-06-15 14:20:30 -07:00
reports
results
reviewer_registry
sample_efficiency
self_consistency_over_time
sensorimotor_sensorium Add sensorium eval and governance runway 2026-06-03 20:53:05 -07:00
sensorium Add event vision sensorium lane 2026-06-06 12:37:57 -07:00
set_membership evals: add staged independent gold lanes 2026-06-05 16:03:26 -07:00
setup_oracle feat(comprehension): inverse reader frame — base of a more/fewer-than (PR-7b / R2 C0) 2026-06-07 07:06:26 -07:00
syllogism feat(comprehend): multi-word NP chunking under a canonicalization contract 2026-06-05 22:47:34 -07:00
symbolic_logic
teaching_injection_resistance
total_ordering feat(comprehend): multi-word NP chunking under a canonicalization contract 2026-06-05 22:47:34 -07:00
vision_sensorium Add vision evidence and sensorimotor contracts 2026-06-03 20:27:46 -07:00
walkthrough_chain
warmed_session_consistency
zero_code_domain_acquisition
__init__.py
_parallel.py
baseline_runner.py
CLAIMS.md
cognition_cases.jsonl
framework.py
holdout_runner.py
metrics.py
parallel.py
run_cognition_eval.py feat(W-008): L10 Shape B hybrid engine-state persistence (#271) 2026-05-25 11:45:54 -07:00