Llama 3.1 70B Instruct Behavioral Fingerprint

How Meta's Llama 3.1 70B Instruct answers one-word probe questions at temperature 1.0 — random numbers, colors, animals, coin flips — measured over 445 valid samples in the open dataset of "One Token Is Enough" (arXiv:2607.10252).

Vendor

Meta

Data generated

2026-07-08

Probe cells

15

Valid samples

445

Randomness score

43 / 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)

53

29 / 30 answers

Favorite color

turquoise

20 / 30 answers

Coin flip: heads

70%

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 = 0.84 bit · uniform baseline 1.00 bit

AnswerCountShare
heads2273.3%
tails826.7%

Coin flip · Chinese

n = 30 · H = 0.92 bit · uniform baseline 1.00 bit

AnswerCountShare
heads2066.7%
tails1033.3%

Favorite number · English

n = 29 · H = 0.69 bit · uniform baseline 13.29 bit

AnswerCountShare
422586.2%
7310.3%
1313.4%

Favorite number · Chinese

n = 29 · H = 3.01 bit · uniform baseline 13.29 bit

AnswerCountShare
71137.9%
42310.3%
88310.3%
026.9%
126.9%
826.9%
313.4%
613.4%

+ 4 more answers

Random animal · English

n = 30 · H = 2.19 bit · uniform baseline 5.64 bit

AnswerCountShare
kangaroo1756.7%
rhinoceros413.3%
otter26.7%
koala26.7%
quokka13.3%
tiger13.3%
mongoose13.3%
leopard13.3%

+ 1 more answers

Random animal · Chinese

n = 27 · H = 3.78 bit · uniform baseline 5.64 bit

AnswerCountShare
狮子414.8%
311.1%
311.1%
大象311.1%
袋鼠27.4%
鸵鸟27.4%
刺猬13.7%
蝙蝠13.7%

+ 8 more answers

Random city · English

n = 30 · H = 3.17 bit · uniform baseline 5.64 bit

AnswerCountShare
tokyo930.0%
omaha723.3%
bangkok26.7%
ottawa26.7%
pittsburgh13.3%
boston13.3%
sydney13.3%
tulsa13.3%

+ 6 more answers

Random city · Chinese

n = 30 · H = 4.42 bit · uniform baseline 5.64 bit

AnswerCountShare
南京310.0%
杭州26.7%
慕尼黑26.7%
广州26.7%
北京26.7%
大阪26.7%
郑州13.3%
大连13.3%

+ 15 more answers

Random color · English

n = 30 · H = 1.48 bit · uniform baseline 4.91 bit

AnswerCountShare
turquoise2066.7%
magenta516.7%
purple310.0%
cyan13.3%
teal13.3%

Random color · Chinese

n = 30 · H = 2.99 bit · uniform baseline 4.91 bit

AnswerCountShare
826.7%
413.3%
413.3%
413.3%
310.0%
绿310.0%
13.3%
13.3%

+ 2 more answers

Random letter · English

n = 30 · H = 0.85 bit · uniform baseline 4.70 bit

AnswerCountShare
k2480.0%
j516.7%
q13.3%

Random number 1-10 · English

n = 30 · H = 0.21 bit · uniform baseline 3.32 bit

AnswerCountShare
72996.7%
813.3%

Random number 1-10 · Chinese

n = 30 · H = 1.24 bit · uniform baseline 3.32 bit

AnswerCountShare
72273.3%
8413.3%
526.7%
626.7%

Random number 1-100 · English

n = 30 · H = 0.21 bit · uniform baseline 6.64 bit

AnswerCountShare
532996.7%
8313.3%

Random number 1-100 · Chinese

n = 30 · H = 3.44 bit · uniform baseline 6.64 bit

AnswerCountShare
53620.0%
67516.7%
52413.3%
57310.0%
5426.7%
8726.7%
4713.3%
5113.3%

+ 6 more answers

Want to verify your API really serves Llama 3.1 70B Instruct?

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 Meta model fingerprints

Browse all 167 model fingerprints

Llama 3.1 70B Instruct Behavioral Fingerprint — Favorite Random Numbers & Distribution | Tosea.ai