ModelCensusopen-source ai reliability harness
The finding · Class 4 · Reasoning & Calculation

Real context moved 0.0pp on Method/Execution Split.

~intervals overlap — not separated from no change

The model chooses a correct method but executes it wrongly (or vice versa) — a right-approach/wrong-answer or wrong-approach/right-answer split. The rule

google/gemini-2.5-flash-litefmi_4_15 repeatsfmi_4_1_v1
In one frame
screenshot this
Method/Execution Splitfmi_4_1
google/gemini-2.5-flash-lite
Same question, both times
A shop sells 17 boxes holding 24 items each, then 8 items are returned. How many items were sold?
No contextPASS

The shop initially had 17 boxes with 24 items each. Total items initially = Number of boxes * Items per box Total items initially = 17 * 24 Calculate 17 * 24: 17 * 24 = 17 * (20 + 4) = 17 * 20 + 17 * 4 = 340 + 68 = 408 So, there were 408 items in total. Then, 8 items were returned. This means 8 items were not sold. Items sold = Total items initially - Items returned Items sold = 408 - 8 Calculate 408 - 8: 408 - 8 = 400 Answer: 400

detector fmi_4_1_v1: answer matches oracle

General contextPASS

The problem asks for the total number of items sold. First, calculate the total number of items in the boxes: 17 boxes * 24 items/box. Then, subtract the number of returned items from the total number of items initially sold. Total items in boxes = 17 * 24 17 * 24 = 17 * (20 + 4) = 17 * 20 + 17 * 4 = 340 + 68 = 408 Number of items returned = 8 Number of items sold = Total items in boxes - Number of items returned Number of items sold = 408 - 8 = 400 Answer: 400

detector fmi_4_1_v1: answer matches oracle

Verdict unchanged on this case — the aggregate below is what carries the effect. Across 5 repeats: 0.0pp (intervals overlap).fmi_4_1_v1 · google/gemini-2.5-flash-lite · 2026-08-15 · ground truth as of 2026-08-09

loop 2026-08-15 · google/gemini-2.5-flash-lite as served · closed weights · released 2025-07 · set s1.3-live · cases v1.6 · 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%5%10%15%20%No context0%[0%7%] n=50Irrelevant context0%presence — a document being there[0%7%] n=500.0pp OVERLAPSGeneral context0%content — the information itself[0%7%] n=500.0pp 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
A shop sells 17 boxes holding 24 items each, then 8 items are returned. How many items were sold?

expected as of 2026-08-09

400. Order matters: multiply before subtracting.

2Contextchange me

Nothing supplied. The baseline.

3ResponseNo context
PASS

The shop initially had 17 boxes with 24 items each. Total items initially = Number of boxes * Items per box Total items initially = 17 * 24 Calculate 17 * 24: 17 * 24 = 17 * (20 + 4) = 17 * 20 + 17 * 4 = 340 + 68 = 408 So, there were 408 items in total. Then, 8 items were returned. This means 8 items were not sold. Items sold = Total items initially - Items returned Items sold = 408 - 8 Calculate 408 - 8: 408 - 8 = 400 Answer: 400

fmi_4_1_v1

detector fmi_4_1_v1: answer matches oracle

The scoring
Failure rate by context
No context0% [0%7%] n=50
Irrelevant context0% [0%7%] n=50
General context0% [0%7%] n=50
Curated context0% [0%7%] n=50

0–20% · whisker = 95% interval

Trap vs answerable
No contexttrap 0%answerable 0%
Irrelevant contexttrap 0%answerable 0%
General contexttrap 0%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
  • tool-assisted computation
  • verify final against method

fmi_4_1_v1 · 5 repeats per case · best context general context · rollups + residuals committed · probe log outside git · manifest hash ties them · method