Hermes 4 405B Behavioral Fingerprint
How Nous's Hermes 4 405B answers one-word probe questions at temperature 1.0 — random numbers, colors, animals, coin flips — measured over 380 valid samples in the open dataset of "One Token Is Enough" (arXiv:2607.10252).
Vendor
Nous
Data generated
2026-07-08
Probe cells
15
Valid samples
380
Randomness score
33 / 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)
1
4 / 12 answers
Favorite color
blue
17 / 21 answers
Coin flip: heads
20%
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 = 0.00 bit · uniform baseline 1.00 bit
| Answer | Count | Share |
|---|---|---|
| tails | 30 | 100.0% |
Coin flip · Chinese
n = 29 · H = 0.98 bit · uniform baseline 1.00 bit
| Answer | Count | Share |
|---|---|---|
| tails | 17 | 58.6% |
| heads | 12 | 41.4% |
Favorite number · English
n = 28 · H = 0.59 bit · uniform baseline 13.29 bit
| Answer | Count | Share |
|---|---|---|
| 7 | 25 | 89.3% |
| 42 | 2 | 7.1% |
| 17 | 1 | 3.6% |
Favorite number · Chinese
n = 30 · H = 0.21 bit · uniform baseline 13.29 bit
| Answer | Count | Share |
|---|---|---|
| 7 | 29 | 96.7% |
| 5 | 1 | 3.3% |
Random animal · English
n = 13 · H = 2.35 bit · uniform baseline 5.64 bit
| Answer | Count | Share |
|---|---|---|
| elephant | 4 | 30.8% |
| kangaroo | 3 | 23.1% |
| giraffe | 3 | 23.1% |
| lion | 1 | 7.7% |
| hyena | 1 | 7.7% |
| aardvark | 1 | 7.7% |
Random animal · Chinese
n = 29 · H = 3.77 bit · uniform baseline 5.64 bit
| Answer | Count | Share |
|---|---|---|
| 狗 | 4 | 13.8% |
| 鸵鸟 | 4 | 13.8% |
| cat | 3 | 10.3% |
| 猫 | 3 | 10.3% |
| 貓 | 2 | 6.9% |
| 河马 | 2 | 6.9% |
| 恐龙 | 2 | 6.9% |
| 袋鼠 | 1 | 3.4% |
+ 8 more answers
Random city · English
n = 21 · H = 3.59 bit · uniform baseline 5.64 bit
| Answer | Count | Share |
|---|---|---|
| paris | 4 | 19.0% |
| portland | 3 | 14.3% |
| barcelona | 2 | 9.5% |
| lisbon | 2 | 9.5% |
| tokyo | 1 | 4.8% |
| seattle | 1 | 4.8% |
| vancouver | 1 | 4.8% |
| philadelphia | 1 | 4.8% |
+ 6 more answers
Random city · Chinese
n = 29 · H = 2.16 bit · uniform baseline 5.64 bit
| Answer | Count | Share |
|---|---|---|
| 上海 | 17 | 58.6% |
| 巴黎 | 3 | 10.3% |
| 伦敦 | 2 | 6.9% |
| 成都 | 2 | 6.9% |
| 悉尼 | 1 | 3.4% |
| 柏林 | 1 | 3.4% |
| sexo | 1 | 3.4% |
| 東京 | 1 | 3.4% |
+ 1 more answers
Random color · English
n = 21 · H = 0.89 bit · uniform baseline 4.91 bit
| Answer | Count | Share |
|---|---|---|
| blue | 17 | 81.0% |
| turquoise | 2 | 9.5% |
| green | 2 | 9.5% |
Random color · Chinese
n = 29 · H = 1.13 bit · uniform baseline 4.91 bit
| Answer | Count | Share |
|---|---|---|
| 蓝 | 22 | 75.9% |
| 绿 | 4 | 13.8% |
| 红 | 2 | 6.9% |
| 蓝绿 | 1 | 3.4% |
Random letter · English
n = 24 · H = 2.90 bit · uniform baseline 4.70 bit
| Answer | Count | Share |
|---|---|---|
| q | 9 | 37.5% |
| k | 4 | 16.7% |
| g | 2 | 8.3% |
| p | 2 | 8.3% |
| m | 1 | 4.2% |
| a | 1 | 4.2% |
| x | 1 | 4.2% |
| n | 1 | 4.2% |
+ 3 more answers
Random number 1-10 · English
n = 27 · H = 0.23 bit · uniform baseline 3.32 bit
| Answer | Count | Share |
|---|---|---|
| 7 | 26 | 96.3% |
| 8 | 1 | 3.7% |
Random number 1-10 · Chinese
n = 30 · H = 0.21 bit · uniform baseline 3.32 bit
| Answer | Count | Share |
|---|---|---|
| 7 | 29 | 96.7% |
| 5 | 1 | 3.3% |
Random number 1-100 · English
n = 12 · H = 2.42 bit · uniform baseline 6.64 bit
| Answer | Count | Share |
|---|---|---|
| 1 | 4 | 33.3% |
| 42 | 2 | 16.7% |
| 57 | 2 | 16.7% |
| 73 | 2 | 16.7% |
| 58 | 1 | 8.3% |
| 78 | 1 | 8.3% |
Random number 1-100 · Chinese
n = 28 · H = 1.51 bit · uniform baseline 6.64 bit
| Answer | Count | Share |
|---|---|---|
| 42 | 20 | 71.4% |
| 37 | 2 | 7.1% |
| 57 | 2 | 7.1% |
| 58 | 2 | 7.1% |
| 47 | 1 | 3.6% |
| 74 | 1 | 3.6% |
Want to verify your API really serves Hermes 4 405B?
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 endpointData 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.