id: the-missing-column-trailer
title: The Missing Column — trailer
cohort: A-derivative
grammar: >-
  four acts from the record, cut for a feed: paper-and-ink construction, the licensed interval, the census
  tallies, one released file
duration_s: 34
thesis: >-
  Two identical scores license a whole interval of joint outcomes; one assumption picks a point inside
  it; of 20 public evaluations none documents matched thresholds with full exposure and 5 carry any joint-evidence
  artifact; on the one released file the marginals leave 13 possible counts and the rows say 9.
epistemic_operation: >-
  fixed marginals with moving overlap → a set of feasible worlds (I–II); classification against frozen
  criteria → mechanical counts (III); marginals → sharp finite identified set → a union identifies the
  count (IV)
claim: >-
  For marginals (0.10, 0.10) the both-fail probability is identified only to [0.0, 0.10] and independence
  names 0.01 (CC-001, CC-004). As of 2026-08-27, of 20 examined evaluations 14 establish a shared basis,
  0 document matched thresholds with full exposure, and 5 provide a joint-evidence artifact, 2 of them
  via a per-item release (MC-001). On the BELLS 2025 released subset the five marginals identify
  the all-miss count only to {0 … 12} of 82; the released rows give 9; the independence plug-in expects
  2.87 (MC-002, MC-003).
scope: >-
  A montage of four bound facts with no new arithmetic: the construction and interval are the Same Scores
  master's (illustrative population, CC-001 input, CC-004 witnesses); the census counts are verify_census.compute_counts
  on census.yaml at the freeze, one primary reviewer, bounded search; the BELLS numbers are counting arithmetic
  on one author-selected file at native, unstated operating points. Nothing on screen is a measured deployed
  stack.
status: >-
  supported_within_scope — bound to CC-001, CC-004, MC-001, MC-002, MC-003 at their registered commits
evidence:
- fact: CC-001.marginals
  kind: REGISTRY
  value:
  - 0.1
  - 0.1
- fact: CC-001.and_bounds
  kind: PROVED
  value:
  - 0.0
  - 0.1
- fact: CC-001.independence_and
  kind: DERIVED
  value: 0.010000000000000002
- fact: CC-004.witness_lower
  kind: PROVED
  value:
  - 0.8
  - 0.1
  - 0.1
  - 0.0
- fact: CC-004.witness_upper
  kind: PROVED
  value:
  - 0.9
  - 0.0
  - 0.0
  - 0.1
- fact: MC-001.N
  kind: DERIVED
  value: 20
- fact: MC-001.M
  kind: DERIVED
  value: 14
- fact: MC-001.M3
  kind: DERIVED
  value: 0
- fact: MC-001.K
  kind: DERIVED
  value: 5
- fact: MC-001.K.releases_computable_items
  kind: DERIVED
  value: 2
- fact: MC-001.frozen_as_of
  kind: REGISTRY
  value: '2026-08-27'
- fact: MC-002.n_harmful
  kind: OBSERVED
  value: 82
- fact: MC-002.all_miss
  kind: OBSERVED
  value: 9
- fact: MC-002.independence_plugin_prompts
  kind: DERIVED
  value: 2.8657
- fact: MC-003.identified_set_lower
  kind: PROVED
  value: 0
- fact: MC-003.identified_set_upper
  kind: PROVED
  value: 12
objects:
- object: the 100-cell sheet and which cells each guard misses
  status: ILLUSTRATIVE
- object: the readouts GUARD A MISSES 10/100 and GUARD B MISSES 10/100, marked FIXED
  status: REGISTRY
  note: the declared input of CC-001; unchanged in every frame
- object: the BOTH MISS counter running 0 → 10
  status: PROVED
  note: >-
    the end states are CC-004 endpoint witnesses; every intermediate state is feasible by the same construction
- object: the interval 0%–10% over the dimmed sheet
  status: PROVED
  note: CC-001 bounds
- object: the amber point at 1% labelled independence picks
  status: DERIVED
  note: the product of the marginals — what one assumption selects, not an answer
- object: the census tallies 20 / 14 / 0 / 5 and the 2-via-release qualifier
  status: DERIVED
  note: verify_census.compute_counts; the 0 is a reporting fact
- object: the BELLS axis {0 … 12} of 82, the pinned 9, the dashed 2.87
  status: PROVED
  note: >-
    the set is MC-003; 9 is MC-002 OBSERVED counting on the released file; 2.87 is DERIVED from the plug-in
- object: the title, the action and the locator
  status: ILLUSTRATIVE
  note: a route, not a claim
evidence_commits:
- claim: CC-001
  repo: Cubits11/cc-framework
  commit: 167aa1ee514ee3d797ed9be2c60120d0c1db71f6
- claim: CC-004
  repo: Cubits11/cc-framework
  commit: 167aa1ee514ee3d797ed9be2c60120d0c1db71f6
- claim: MC-001
  repo: Cubits11/cubits11.github.io
  commit: 6a6f587a2c0a5dbd1e8fea065c383938877fe3b0
- claim: MC-002
  repo: CentreSecuriteIA/bells_leaderboard
  commit: 507566c5a4606c8e3dec0bd59a5c5fde62594951
- claim: MC-003
  repo: Cubits11/cubits11.github.io
  commit: 6a6f587a2c0a5dbd1e8fea065c383938877fe3b0
falsifier: >-
  A clean-clone execution of the bound kernel on marginals (0.10, 0.10) returns AND bounds other than
  [0.0, 0.10] (CC-001/CC-004, REJECT); scripts/verify_census.py --counts prints counts other than 20 examined, 14 with a shared basis, 0 with matched thresholds and full exposure, 5 with a joint-evidence artifact, and 2 with per-item releases (MC-001, REJECT); scripts/reanalyze_bells_subset.py recomputes an all-miss
  other than 9/82 or scripts/identification.py a set other than {0 … 12} (MC-002/MC-003, REJECT); or any
  rendered frame shows a readout other than 10/100 for either guard while the BOTH counter moves — in
  each case bind_facts.py --check fails before this film can be re-rendered.
non_claims:
- does not certify any stacked system as safe and shows no measured deployed stack
- >-
  does not estimate or recover the true dependence between any two guards; the amber points are what one
  assumption would select
- >-
  does not claim any evaluated guardrail performs poorly; an unfilled column is a reporting fact
- >-
  does not treat the BELLS 9/82 as a rate or a population estimate; it is counting arithmetic on one author-selected
  released file
- the census is a bounded single-reviewer search, not proof of universal absence
# carried verbatim from claims.yaml (S3 of spine.yaml): MC-003, MC-002, MC-001, CC-001
- "no new mathematics is claimed; the bounds are Fréchet's and the lower endpoint is Bonferroni's"
- "not a claim that guardrails in general fail together — this stratum cannot separate shared blind spots from prompt-difficulty heterogeneity"
- "not a vendor evaluation; the released verdicts are at unstated default configurations, and one supervisor fires exactly once in the 170 released rows"
- "the identified interval is what the marginals leave open, not a prediction about any deployed stack"
- "leave-one-out unions identify only exclusive full-stack coverage; they do not identify pairwise or higher-order overlap, Shapley values, or causal contribution"
- "says nothing about adversarial prompts, which dominate the full evaluation and have no per-item release"
- "the subset is author-selected with an unstated rule — nothing here estimates any system's true rate, and no confidence interval is offered because the sampled population is undefined"
- "not a ranking or endorsement of any vendor; the released verdicts are at unstated default configurations"
- "says nothing about adversarial prompts, which dominate the full evaluation and have no per-item release"
- "the release-recomputed-to-plug-in ratio describes this subset's arithmetic, not a general law of guardrail dependence"
- "not an observed deployed five-guard stack; interpreting this static OR aggregation as a stack requires the separate full-exposure, parallel, fixed-operating-point assumptions"
- "does not claim any evaluated guardrail stack performs poorly; an unfilled column is a reporting fact, not a performance finding"
- "does not claim the unmeasured joint statistics would reveal dependence; measuring instead of assuming is the point"
- "does not audit the quality of any per-system evaluation beyond the fields each row records"
- "does not claim exhaustive coverage beyond the corrected, examined artifact set or coverage after the stated date"
- "does not certify any stacked system as safe"
- "does not recover or estimate the unknown dependence"
- "does not select a point inside the returned interval"
claim_frames:
- t: 0.4
  shows: opening — A 10/100, B 10/100, both 0/100, nothing has moved
- t: 3.0
  shows: mid-transition — half the rings landed, readouts unchanged
- t: 7.0
  shows: opposite endpoint 10/100 and the name SAME SCORES. DIFFERENT WORLDS.
- t: 12.0
  shows: the interval 0–10% with the amber 1% labelled as what independence picks
- t: 20.0
  shows: the census — 20 · 14 · 0 · 5 (2 via release), frozen 2026-08-27
- t: 27.5
  shows: BELLS — the set {0 … 12} of 82, the released 9, the amber 2.87 expectation
- t: 32.4
  shows: the close — title, BRING THE COUNTEREXAMPLE, locator
render:
  kind: master
  fps: 30
  formats:
  - master
  poster_t: 32.4
cta_route: /missing-column/