The finding · Class 1 · Grounding & Attribution

Real context moved -1.2pp on Citation Resolution Failure.

~intervals overlap — not separated from no change

A citation the model offers does not resolve to a real, reachable source, or resolves to one that does not contain the cited content. The rule

anthropic/claude-3-haikufmi_1_25 repeatsfmi_1_2_v1

The mode All 3 cards

loop 2026-08-15 · anthropic/claude-3-haiku as served · closed weights · released 2024-03 · set s1.1-live · cases v1.3 · conditions v1.0 · ground truth as of 2026-08-09

From no context to real context

5 repeats per case

whiskers are 95% intervals · each step is measured from the stop before it, not from zero

Try the other contexts

stored · no model call

No exhibits stored for this card.

The scoring

Failure rate by context

No context70% [48%–85%] n=20

Irrelevant context60% [39%–78%] n=20

General context59% [36%–78%] n=17

Curated context70% [48%–85%] n=20

0–100% · whisker = 95% interval

Trap vs answerable

No contexttrap 90%answerable 50%

Irrelevant contexttrap 70%answerable 50%

General contexttrap 71%answerable 50%

Curated contexttrap 90%answerable 50%

never pooled · a trap ratio would skew the delta

Residual — what survives

70% with no context, 59% under the best supplied source. What survives is 10 failing probes at that best condition, the detector's reason on most of them being: no citation offered.

Mitigations from the index — the claim

post-hoc citation verification
retrieval-then-cite
refuse to fabricate identifiers

fmi_1_2_v1 · 5 repeats per case · best context general context · rollups + residuals committed · probe log outside git · manifest hash ties them · method