Real context moved 0.0pp on Paraphrase Non-Invariance.
~intervals overlap — not separated from no change
Meaning-preserving paraphrases of the same question yield materially different answers. The rule
How many bones are in the adult human body?
The adult human skeleton typically consists of **206 bones**. This total is conventionally divided as follows: - Axial skeleton: 80 bones - Appendicular skeleton: 126 bones (Note that infants are born with ~270 bones; many of these fuse during growth and development to produce the adult count. Minor individual variation can occur due to accessory/sesamoid bones, but 206 remains the standard figure.)
detector fmi_5_2_v1: answers invariant across paraphrases
**206** An adult human skeleton typically has 206 bones (after fusion of many of the ~270 bones present at birth). Minor individual variation can occur (e.g., extra sesamoid bones or unfused sutures), but 206 is the standard count.
detector fmi_5_2_v1: answers invariant across paraphrases
loop 2026-08-15 · x-ai/grok-4.6 as served · closed weights · released 2026-01 · set s1.3-live · cases v1.6 · conditions v1.0 · ground truth as of 2026-08-09
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.
Nothing supplied. The baseline.
The adult human skeleton typically consists of **206 bones**. This total is conventionally divided as follows: - Axial skeleton: 80 bones - Appendicular skeleton: 126 bones (Note that infants are born with ~270 bones; many of these fuse during growth and development to produce the adult count. Minor individual variation can occur due to accessory/sesamoid bones, but 206 remains the standard figure.)
fmi_5_2_v1
detector fmi_5_2_v1: answers invariant across paraphrases
0–20% · whisker = 95% interval
never pooled · a trap ratio would skew the delta
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.
- 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