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

Real context moved -7.5pp on Date-Arithmetic Error.

~intervals overlap — not separated from no change

Given an injected "current date," the model computes relative dates/intervals incorrectly. The rule

anthropic/claude-sonnet-4.6fmi_7_35 repeatsfmi_7_3_v1
The modeAll 7 cards
In one frame
screenshot this
Date-Arithmetic Errorfmi_7_3
anthropic/claude-sonnet-4.6
Same question, both times
Today is 2026-01-30. What date is 30 days from today? Answer with the full date.
No contextPASS

Today's date is **2026-01-30**. Adding 30 days: - January has 31 days, so from January 30, there is 1 day remaining in January (January 31). - That leaves 29 more days into February. - February 2026 has 28 days (2026 is not a leap year), so 29 days into February goes 1 day into March. **March 1, 2026**

detector fmi_7_3_v1: date arithmetic correct

Curated contextPASS

30 days from January 30, 2026 is **March 1, 2026**.

detector fmi_7_3_v1: date arithmetic correct

Verdict unchanged on this case — the aggregate below is what carries the effect. Across 5 repeats: -7.5pp (intervals overlap).fmi_7_3_v1 · anthropic/claude-sonnet-4.6 · 2026-08-15 · ground truth as of 2026-08-09

loop 2026-08-15 · anthropic/claude-sonnet-4.6 as served · closed weights · released 2026-02 · set s1.2-live · cases v1.4 · 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%10%20%30%40%No context0%[0%–9%] n=40Irrelevant context8%presence — a document being there[3%–20%] n=40+7.5pp OVERLAPSCurated context0%content — the information itself[0%–9%] n=40-7.5pp 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
Today is 2026-01-30. What date is 30 days from today? Answer with the full date.

expected as of 2026-08-09

From 2026-01-30, 30 days later is 2026-03-01 (1 March 2026). Crossing February in a non-leap year.

2Contextchange me

Nothing supplied. The baseline.

3ResponseNo context
PASS

Today's date is **2026-01-30**. Adding 30 days: - January has 31 days, so from January 30, there is 1 day remaining in January (January 31). - That leaves 29 more days into February. - February 2026 has 28 days (2026 is not a leap year), so 29 days into February goes 1 day into March. **March 1, 2026**

fmi_7_3_v1

detector fmi_7_3_v1: date arithmetic correct

The scoring
Failure rate by context
No context0% [0%–9%] n=40
Irrelevant context8% [3%–20%] n=40
General context3% [0%–13%] n=40
Curated context0% [0%–9%] n=40

0–40% · whisker = 95% interval

Trap vs answerable
No contexttrap 0%answerable 0%
Irrelevant contexttrap 12%answerable 0%
General contexttrap 4%answerable 0%
Curated contexttrap 0%answerable 0%

never pooled · a trap ratio would skew the delta

Residual — what survives

Unchanged by context: 0% with none, 0% with the strongest supplied source. For a control mode that flat line is the result — it is what licenses reading movement elsewhere in this loop as grounding rather than as a document being present. Nothing survives at the best condition on the cases measured here.

Mitigations from the index — the claim
  • date libraries via tools
  • explicit date arithmetic

fmi_7_3_v1 · 5 repeats per case · best context curated context · rollups + residuals committed · probe log outside git · manifest hash ties them · method