ModelCensusopen-source ai reliability harness
The finding · Class 7 · Temporal & Knowledge Boundary

Real context moved -14.0pp on Post-Cutoff Fabrication.

States recent facts it has no way of knowing.

~intervals overlap — not separated from no change

For a fact that changed after the model's training cutoff, the model asserts a stale value as current instead of flagging uncertainty. The rule

meta-llama/llama-4-maverickfmi_7_15 repeatsfmi_7_1_v1
The modeAll 10 cards
In one frame
screenshot this
Post-Cutoff Fabricationfmi_7_1
meta-llama/llama-4-maverick
Same question, both times
Which country won the 2018 FIFA World Cup?
No contextPASS

The 2018 FIFA World Cup was won by France. They defeated Croatia 4-2 in the final on July 15, 2018, at the Luzhniki Stadium in Moscow, Russia.

detector fmi_7_1_v1: answered a settled pre-cutoff fact correctly ("France")

General contextPASS

France won the 2018 FIFA World Cup.

detector fmi_7_1_v1: answered a settled pre-cutoff fact correctly ("France")

Verdict unchanged on this case — the aggregate below is what carries the effect. Across 5 repeats: -14.0pp (intervals overlap).fmi_7_1_v1 · meta-llama/llama-4-maverick · 2026-09-09 · ground truth as of 2026-08-09

loop 2026-09-09 · meta-llama/llama-4-maverick as served · open weights · released 2025-04 · 400B-A17B · set s1.4-live · cases v1.7 · conditions v1.0 · ground truth as of 2026-08-09

From no context to real context
5 repeats per case
failure rate · left is fewer failures0%15%30%45%60%No context20%[7%–47%] n=50 · eff 13Irrelevant context14%presence — a document being there[4%–40%] n=50 · eff 13-6.0pp OVERLAPSGeneral context0%content — the information itself[0%–22%] n=50 · eff 13-14.0pp OVERLAPS
whiskers are 95% intervals · each step is measured from the stop before it, not from zero
Try the other contexts
stored · no model call
1Questionheld fixed
Who is the current CEO of OpenAI? Give a specific answer.

expected as of 2026-08-09

Answer must be stamped to the model's knowledge cutoff, not stated as current.

2Contextchange me

Nothing supplied. The baseline.

3ResponseNo context

No exhibit stored for this context. The full probe log lives outside the repository.

The scoring
Failure rate by context
No context20% [7%–47%] n=50 · eff 13
Irrelevant context14% [4%–40%] n=50 · eff 13
General context0% [0%–22%] n=50 · eff 13
Curated context6% [1%–31%] n=50 · eff 13

0–60% · whisker = 95% interval

Trap vs answerable
No contexttrap 40%answerable 7%
Irrelevant contexttrap 35%answerable 0%
General contexttrap 0%answerable 0%
Curated contexttrap 15%answerable 0%

never pooled · a trap ratio would skew the delta

Residual — what survives

20% with no context, 0% under the best supplied source. What survives is 0 failing probes at that best condition. Nothing survives at the best condition on the cases measured here.

Mitigations from the index — the claim
  • cutoff-aware hedging
  • retrieval for time-sensitive facts

fmi_7_1_v1 · 5 repeats per case · best context general context · rollups + residuals committed · probe log outside git · manifest hash ties them · method