ModelCensusopen-source ai reliability harness
The finding · Class 5 · Robustness & Consistency

Real context moved 3.3pp on Paraphrase Non-Invariance.

Same question, reworded, different answer.

Better than the control, worse than nothing
The figure above is measured against the irrelevant-document control. Against no context at all this card is worse: 0% bare against 3% with the best supplied source. What the source repaired is damage the presence of a document did.

~intervals overlap — not separated from no change

Meaning-preserving paraphrases of the same question yield materially different answers. The rule

google/gemma-3-27b-itfmi_5_25 repeatsfmi_5_2_v1
In one frame
screenshot this
Paraphrase Non-Invariancefmi_5_2
google/gemma-3-27b-it
Same question, both times
How many bones are in the adult human body?
No contextPASS

The total number of bones in an adult human is **206**. However, this is a bit of a nuanced answer! Here's a breakdown: * **Babies are born with around 300 bones.** These include many smaller bones that fuse together as they grow. * **Fusion happens over time.** As children grow, many separate bones gradually fuse (join together) to form larger, stronger bones. * **206 is the number after fusion is largely complete.** This count doesn't include small sesamoid bones (like the patella/kneecap, which can vary in number) or tiny ossicles in the ears. **Here's a general breakdown of how the 206 bones are categorized:** * **Skull:** 22 bones * **Spine (Vertebrae):** 26 bones * **Ribs:** 24 bones (12 pairs) * **Appendages (Arms & Legs):** 64 bones (32 in each arm/leg) * **Hands & Feet:** 52 bones (27 in each hand/foot) * **Pelvis:** 2 bones (hip bones) * **Sternum:** 1 bone * **Hyoid:** 1 bone (in the neck)

detector fmi_5_2_v1: answers invariant across paraphrases

General contextPASS

This passage discusses regional variations in English vocabulary for a type of bread roll. It **does not contain information about the number of bones in an adult human**. The answer to your question is **206**. However, it is not found in the provided text.

detector fmi_5_2_v1: answers invariant across paraphrases

Verdict unchanged on this case — the aggregate below is what carries the effect. Across 5 repeats: +3.3pp (intervals overlap).fmi_5_2_v1 · google/gemma-3-27b-it · 2026-09-09 · ground truth as of 2026-08-09

loop 2026-09-09 · google/gemma-3-27b-it as served · open weights · released 2025-03 · 27B · 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 context0%[0%34%] n=30 · eff 7Irrelevant context0%presence — a document being there[0%34%] n=30 · eff 70.0pp OVERLAPSGeneral context3%content — the information itself[0%39%] n=30 · eff 7+3.3pp 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
How many bones are in the adult human body?

expected as of 2026-08-09

206. 206 in a typical adult. Widely misremembered, which is what makes the phrasings diverge.

2Contextchange me

Nothing supplied. The baseline.

3ResponseNo context
PASS

The total number of bones in an adult human is **206**. However, this is a bit of a nuanced answer! Here's a breakdown: * **Babies are born with around 300 bones.** These include many smaller bones that fuse together as they grow. * **Fusion happens over time.** As children grow, many separate bones gradually fuse (join together) to form larger, stronger bones. * **206 is the number after fusion is largely complete.** This count doesn't include small sesamoid bones (like the patella/kneecap, which can vary in number) or tiny ossicles in the ears. **Here's a general breakdown of how the 206 bones are categorized:** * **Skull:** 22 bones * **Spine (Vertebrae):** 26 bones * **Ribs:** 24 bones (12 pairs) * **Appendages (Arms & Legs):** 64 bones (32 in each arm/leg) * **Hands & Feet:** 52 bones (27 in each hand/foot) * **Pelvis:** 2 bones (hip bones) * **Sternum:** 1 bone * **Hyoid:** 1 bone (in the neck)

fmi_5_2_v1

detector fmi_5_2_v1: answers invariant across paraphrases

The scoring
Failure rate by context
No context0% [0%34%] n=30 · eff 7
Irrelevant context0% [0%34%] n=30 · eff 7
General context3% [0%39%] n=30 · eff 7
Curated context7% [1%43%] n=30 · eff 7

0–60% · whisker = 95% interval

Trap vs answerable
No contexttrap 0%answerable 0%
Irrelevant contexttrap 0%answerable 0%
General contexttrap 10%answerable 0%
Curated contexttrap 20%answerable 0%

never pooled · a trap ratio would skew the delta

Residual — what survives

Supplying context did not reduce this failure, it increased it: 0% with no context against 7% under the strongest supplied source. What survives is therefore not a remainder but a substitution — the failures under context are 1 probes the detector still rejected, dominated by: answers differ across meaning-preserving paraphrases.

Mitigations from the index — the claim
  • paraphrase-augmented training
  • canonicalization

fmi_5_2_v1 · 5 repeats per case · best context general context · rollups + residuals committed · probe log outside git · manifest hash ties them · method