Saba Behavioral Fingerprint

How Mistral's Saba answers one-word probe questions at temperature 1.0 — random numbers, colors, animals, coin flips — measured over 449 valid samples in the open dataset of "One Token Is Enough" (arXiv:2607.10252).

Vendor

Mistral

Data generated

2026-07-08

Probe cells

15

Valid samples

449

Randomness score

44 / 100

Mean normalized entropy across cells, 0–100. 100 would be a uniform-random baseline; real models score far lower.

Signature answers

Favorite random number (1–100)

42

16 / 30 answers

Favorite color

red

10 / 30 answers

Coin flip: heads

72%

Share of heads across coin-flip cells.

Answer distributions by probe cell

Empirical distributions of normalized one-word answers, per task and language. H is the Shannon entropy in bits; the uniform baseline is log2 of the answer-domain size.

Coin flip · English

n = 30 · H = 1.00 bit · uniform baseline 1.00 bit

AnswerCountShare
heads1653.3%
tails1446.7%

Coin flip · Chinese

n = 30 · H = 0.47 bit · uniform baseline 1.00 bit

AnswerCountShare
heads2790.0%
tails310.0%

Favorite number · English

n = 30 · H = 1.07 bit · uniform baseline 13.29 bit

AnswerCountShare
422066.7%
7930.0%
113.3%

Favorite number · Chinese

n = 30 · H = 0.63 bit · uniform baseline 13.29 bit

AnswerCountShare
72790.0%
313.3%
413.3%
513.3%

Random animal · English

n = 30 · H = 2.48 bit · uniform baseline 5.64 bit

AnswerCountShare
giraffe1653.3%
elephant310.0%
raccoon26.7%
lion26.7%
penguin13.3%
frog13.3%
bear13.3%
platypus13.3%

+ 3 more answers

Random animal · Chinese

n = 30 · H = 3.42 bit · uniform baseline 5.64 bit

AnswerCountShare
老虎620.0%
长颈鹿413.3%
袋鼠310.0%
兔子310.0%
企鹅310.0%
310.0%
26.7%
斑马13.3%

+ 5 more answers

Random city · English

n = 30 · H = 1.31 bit · uniform baseline 5.64 bit

AnswerCountShare
tokyo2376.7%
toronto26.7%
london26.7%
berlin13.3%
dublin13.3%
paris13.3%

Random city · Chinese

n = 29 · H = 2.56 bit · uniform baseline 5.64 bit

AnswerCountShare
上海1241.4%
北京620.7%
东京413.8%
伦敦13.4%
长春13.4%
纽约13.4%
广州13.4%
温哥华13.4%

+ 2 more answers

Random color · English

n = 30 · H = 2.58 bit · uniform baseline 4.91 bit

AnswerCountShare
red1033.3%
blue930.0%
green413.3%
pink13.3%
lavender13.3%
turquoise13.3%
chartreuse13.3%
yellow13.3%

+ 2 more answers

Random color · Chinese

n = 30 · H = 2.40 bit · uniform baseline 4.91 bit

AnswerCountShare
绿826.7%
723.3%
620.0%
青绿413.3%
413.3%
靛蓝13.3%

Random letter · English

n = 30 · H = 3.77 bit · uniform baseline 4.70 bit

AnswerCountShare
z516.7%
g516.7%
q310.0%
s26.7%
k26.7%
d26.7%
h13.3%
b13.3%

+ 9 more answers

Random number 1-10 · English

n = 30 · H = 0.88 bit · uniform baseline 3.32 bit

AnswerCountShare
72583.3%
5310.0%
313.3%
613.3%

Random number 1-10 · Chinese

n = 30 · H = 1.63 bit · uniform baseline 3.32 bit

AnswerCountShare
72066.7%
3413.3%
426.7%
626.7%
513.3%
813.3%

Random number 1-100 · English

n = 30 · H = 1.92 bit · uniform baseline 6.64 bit

AnswerCountShare
421653.3%
47620.0%
73413.3%
3726.7%
6313.3%
6713.3%

Random number 1-100 · Chinese

n = 30 · H = 2.99 bit · uniform baseline 6.64 bit

AnswerCountShare
42930.0%
47620.0%
67310.0%
73310.0%
1526.7%
8726.7%
1713.3%
2713.3%

+ 3 more answers

Want to verify your API really serves Saba?

Point the free fingerprint checker at your endpoint: it samples the same probe questions from your browser and compares the distributions against this reference. Your API key never leaves your browser.

Verify your endpoint

Data source & license

Distributions: Tomáš Bruckner, "One Token Is Enough" (arXiv:2607.10252), official dataset Zenodo DOI 10.5281/zenodo.21278557, licensed CC-BY-4.0. Collected via OpenRouter at temperature 1.0 under the paper's fixed minimal one-word system prompt; answer keys are re-normalized with the checker's pipeline (color aliases, number words, coin h/t) so live probes are directly comparable.

Paper on arXivDataset on ZenodoCC-BY-4.0 license

More Mistral model fingerprints

Browse all 167 model fingerprints

Saba Behavioral Fingerprint — Favorite Random Numbers & Distribution | Tosea.ai