Grok 4.20 Behavioral Fingerprint
How xAI's Grok 4.20 answers one-word probe questions at temperature 1.0 — random numbers, colors, animals, coin flips — measured over 450 valid samples in the open dataset of "One Token Is Enough" (arXiv:2607.10252).
Vendor
xAI
Data generated
2026-07-08
Probe cells
15
Valid samples
450
Randomness score
24 / 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
blue
30 / 30 answers
Coin flip: heads
97%
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 = 0.35 bit · uniform baseline 1.00 bit
| Answer | Count | Share |
|---|---|---|
| heads | 28 | 93.3% |
| tails | 2 | 6.7% |
Coin flip · Chinese
n = 30 · H = 0.00 bit · uniform baseline 1.00 bit
| Answer | Count | Share |
|---|---|---|
| heads | 30 | 100.0% |
Favorite number · English
n = 30 · H = 0.00 bit · uniform baseline 13.29 bit
| Answer | Count | Share |
|---|---|---|
| 7 | 30 | 100.0% |
Favorite number · Chinese
n = 30 · H = 0.00 bit · uniform baseline 13.29 bit
| Answer | Count | Share |
|---|---|---|
| 7 | 30 | 100.0% |
Random animal · English
n = 30 · H = 2.28 bit · uniform baseline 5.64 bit
| Answer | Count | Share |
|---|---|---|
| elephant | 16 | 53.3% |
| lion | 4 | 13.3% |
| cat | 3 | 10.0% |
| dog | 2 | 6.7% |
| kangaroo | 1 | 3.3% |
| tiger | 1 | 3.3% |
| quokka | 1 | 3.3% |
| aardvark | 1 | 3.3% |
+ 1 more answers
Random animal · Chinese
n = 30 · H = 2.38 bit · uniform baseline 5.64 bit
| Answer | Count | Share |
|---|---|---|
| 狗 | 11 | 36.7% |
| 猫 | 8 | 26.7% |
| 狮子 | 5 | 16.7% |
| 老虎 | 2 | 6.7% |
| 狐狸 | 1 | 3.3% |
| 兔子 | 1 | 3.3% |
| 斑马 | 1 | 3.3% |
| 袋鼠 | 1 | 3.3% |
Random city · English
n = 30 · H = 2.50 bit · uniform baseline 5.64 bit
| Answer | Count | Share |
|---|---|---|
| tokyo | 15 | 50.0% |
| paris | 5 | 16.7% |
| toronto | 2 | 6.7% |
| berlin | 1 | 3.3% |
| phoenix | 1 | 3.3% |
| stockholm | 1 | 3.3% |
| vancouver | 1 | 3.3% |
| new | 1 | 3.3% |
+ 3 more answers
Random city · Chinese
n = 30 · H = 2.32 bit · uniform baseline 5.64 bit
| Answer | Count | Share |
|---|---|---|
| 东京 | 9 | 30.0% |
| tokyo | 8 | 26.7% |
| 上海 | 5 | 16.7% |
| 北京 | 5 | 16.7% |
| shanghai | 2 | 6.7% |
| 巴黎 | 1 | 3.3% |
Random color · English
n = 30 · H = 0.00 bit · uniform baseline 4.91 bit
| Answer | Count | Share |
|---|---|---|
| blue | 30 | 100.0% |
Random color · Chinese
n = 30 · H = 1.46 bit · uniform baseline 4.91 bit
| Answer | Count | Share |
|---|---|---|
| 蓝 | 16 | 53.3% |
| 红 | 7 | 23.3% |
| 紫 | 7 | 23.3% |
Random letter · English
n = 30 · H = 1.82 bit · uniform baseline 4.70 bit
| Answer | Count | Share |
|---|---|---|
| q | 16 | 53.3% |
| z | 8 | 26.7% |
| x | 3 | 10.0% |
| g | 1 | 3.3% |
| k | 1 | 3.3% |
| f | 1 | 3.3% |
Random number 1-10 · English
n = 30 · H = 0.57 bit · uniform baseline 3.32 bit
| Answer | Count | Share |
|---|---|---|
| 7 | 26 | 86.7% |
| 5 | 4 | 13.3% |
Random number 1-10 · Chinese
n = 30 · H = 0.77 bit · uniform baseline 3.32 bit
| Answer | Count | Share |
|---|---|---|
| 7 | 25 | 83.3% |
| 5 | 4 | 13.3% |
| 3 | 1 | 3.3% |
Random number 1-100 · English
n = 30 · H = 1.00 bit · uniform baseline 6.64 bit
| Answer | Count | Share |
|---|---|---|
| 42 | 16 | 53.3% |
| 47 | 14 | 46.7% |
Random number 1-100 · Chinese
n = 30 · H = 1.70 bit · uniform baseline 6.64 bit
| Answer | Count | Share |
|---|---|---|
| 42 | 14 | 46.7% |
| 47 | 12 | 40.0% |
| 63 | 1 | 3.3% |
| 67 | 1 | 3.3% |
| 74 | 1 | 3.3% |
| 83 | 1 | 3.3% |
Want to verify your API really serves Grok 4.20?
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.