cohere/command-r7b-12-2024 behavioral LLM fingerprint
This page breaks down the exact reference that VerifyLLMAPI uses for cohere/command-r7b-12-2024: 118 normalized answers across eight low-entropy probes, plus its nearest enrolled models by mean Jensen–Shannon divergence.
Fingerprint summary
- Reference answers
- 118
- Samples per cell
- 14–15
- Nearest model
- openai/gpt-4o
- Nearest mean JSD
- 0.700650
The most concentrated cell was color-random|zh: 蓝色 held 53.3% of its answers and entropy measured 2.174 bits. The broadest cell was letter-random|zh at 3.774 bits.
Against the closest enrolled model, animal-random|ar contributed the largest cell distance at JSD 1.000000. This is a descriptive comparison, not proof that either API route serves a named checkpoint.
Completeness: the author dataset allows up to 15 valid answers per selected cell in reference half A. This model contains 118/120 across the eight cells. We show the actual denominator for every share.
The eight test requests
The research pipeline sent each user prompt with a one-answer system constraint, temperature 1, a 16-token output cap, and reasoning disabled. VerifyLLMAPI uses the same prompt text and normalization target.
| Cell | Task | Exact request text | Normalization and n |
|---|---|---|---|
animal-random|en |
Random animal · English | Name a random animal.System constraintAnswer with a single word or a single number only. Output nothing else: no explanation, no punctuation, no quotation marks. |
word · 14/15 reference samples |
word-random|en |
Random word · English | Name a random word.System constraintAnswer with a single word or a single number only. Output nothing else: no explanation, no punctuation, no quotation marks. |
word · 14/15 reference samples |
animal-random|ru |
Random animal · Russian | Назови случайное животное.System constraintОтвечай только одним словом или одним числом. Не выводи ничего больше: без объяснений, без знаков препинания, без кавычек. |
word · 15/15 reference samples |
animal-random|ar |
Random animal · Arabic | اذكر حيواناً عشوائياً.System constraintأجب بكلمة واحدة أو رقم واحد فقط. لا تكتب أي شيء آخر: لا شرح ولا علامات ترقيم ولا علامات اقتباس. |
word · 15/15 reference samples |
letter-random|zh |
Random character · Chinese | 说出一个随机的汉字。System constraint只用一个词或一个数字回答。不要输出任何其他内容：不要解释，不要标点符号，不要引号。 |
grapheme · 15/15 reference samples |
animal-random|zh |
Random animal · Chinese | 说出一个随机的动物。System constraint只用一个词或一个数字回答。不要输出任何其他内容：不要解释，不要标点符号，不要引号。 |
word · 15/15 reference samples |
letter-random|ar |
Random letter · Arabic | اذكر حرفاً عشوائياً من الحروف الأبجدية.System constraintأجب بكلمة واحدة أو رقم واحد فقط. لا تكتب أي شيء آخر: لا شرح ولا علامات ترقيم ولا علامات اقتباس. |
grapheme · 15/15 reference samples |
color-random|zh |
Random color · Chinese | 说出一个随机的颜色。System constraint只用一个词或一个数字回答。不要输出任何其他内容：不要解释，不要标点符号，不要引号。 |
word · 15/15 reference samples |
Response distributions
These are normalized reference-half counts from the published OpenRouter dataset. The tables contain aggregate values, not user data or a live VerifyLLMAPI scan.
1. animal-random|en
Random animal, English. 14 valid reference answers produced 13 normalized values. The mode was a at 2/14 (14.3%); entropy was 3.664 bits.
| Normalized response | Count | Share |
|---|---|---|
a |
2 | 14.3% |
arctic |
1 | 7.1% |
bear |
1 | 7.1% |
blue |
1 | 7.1% |
elephant |
1 | 7.1% |
gibbon |
1 | 7.1% |
giraffe |
1 | 7.1% |
jeffrey |
1 | 7.1% |
koala |
1 | 7.1% |
lion |
1 | 7.1% |
penguin |
1 | 7.1% |
tiger |
1 | 7.1% |
الاجتماع |
1 | 7.1% |
2. word-random|en
Random word, English. 14 valid reference answers produced 12 normalized values. The mode was random at 2/14 (14.3%); entropy was 3.522 bits.
| Normalized response | Count | Share |
|---|---|---|
random |
2 | 14.3% |
serendipity |
2 | 14.3% |
edge |
1 | 7.1% |
hector |
1 | 7.1% |
heist |
1 | 7.1% |
july |
1 | 7.1% |
m |
1 | 7.1% |
mystery |
1 | 7.1% |
quixotic |
1 | 7.1% |
sartorial |
1 | 7.1% |
strawberry |
1 | 7.1% |
under |
1 | 7.1% |
3. animal-random|ru
Random animal, Russian. 15 valid reference answers produced 10 normalized values. The mode was лис at 5/15 (33.3%); entropy was 3.000 bits.
| Normalized response | Count | Share |
|---|---|---|
лис |
5 | 33.3% |
л |
2 | 13.3% |
волк |
1 | 6.7% |
лемур |
1 | 6.7% |
лисица |
1 | 6.7% |
лисы |
1 | 6.7% |
лопатая所使用cino |
1 | 6.7% |
львь |
1 | 6.7% |
львьё |
1 | 6.7% |
лёся |
1 | 6.7% |
4. animal-random|ar
Random animal, Arabic. 15 valid reference answers produced 10 normalized values. The mode was الفيل at 4/15 (26.7%); entropy was 3.107 bits.
| Normalized response | Count | Share |
|---|---|---|
الفيل |
4 | 26.7% |
الأسد |
2 | 13.3% |
الدببة |
2 | 13.3% |
ال |
1 | 6.7% |
البارز |
1 | 6.7% |
البطة |
1 | 6.7% |
الفهد |
1 | 6.7% |
الكلب |
1 | 6.7% |
فأر |
1 | 6.7% |
老虎 |
1 | 6.7% |
5. letter-random|zh
Random character, Chinese. 15 valid reference answers produced 14 normalized values. The mode was 秋 at 2/15 (13.3%); entropy was 3.774 bits.
| Normalized response | Count | Share |
|---|---|---|
秋 |
2 | 13.3% |
надц |
1 | 6.7% |
善 |
1 | 6.7% |
微 |
1 | 6.7% |
晚 |
1 | 6.7% |
橙 |
1 | 6.7% |
热 |
1 | 6.7% |
球 |
1 | 6.7% |
琴 |
1 | 6.7% |
米 |
1 | 6.7% |
绿 |
1 | 6.7% |
艺 |
1 | 6.7% |
蛋 |
1 | 6.7% |
魂 |
1 | 6.7% |
6. animal-random|zh
Random animal, Chinese. 15 valid reference answers produced 12 normalized values. The mode was 猫 at 3/15 (20.0%); entropy was 3.457 bits.
| Normalized response | Count | Share |
|---|---|---|
猫 |
3 | 20.0% |
海豚 |
2 | 13.3% |
九龍象鱼 |
1 | 6.7% |
狐狸 |
1 | 6.7% |
狮子 |
1 | 6.7% |
老虎 |
1 | 6.7% |
豚鼠 |
1 | 6.7% |
象 |
1 | 6.7% |
青蛙 |
1 | 6.7% |
驼鹿 |
1 | 6.7% |
鳄鱼 |
1 | 6.7% |
鸵鸟 |
1 | 6.7% |
7. letter-random|ar
Random letter, Arabic. 15 valid reference answers produced 11 normalized values. The mode was س at 4/15 (26.7%); entropy was 3.240 bits.
| Normalized response | Count | Share |
|---|---|---|
س |
4 | 26.7% |
ب |
2 | 13.3% |
ج |
1 | 6.7% |
ح |
1 | 6.7% |
د |
1 | 6.7% |
ز |
1 | 6.7% |
ص |
1 | 6.7% |
ق |
1 | 6.7% |
م |
1 | 6.7% |
و |
1 | 6.7% |
ي |
1 | 6.7% |
8. color-random|zh
Random color, Chinese. 15 valid reference answers produced 7 normalized values. The mode was 蓝色 at 8/15 (53.3%); entropy was 2.174 bits.
| Normalized response | Count | Share |
|---|---|---|
蓝色 |
8 | 53.3% |
橙色 |
2 | 13.3% |
奶油色 |
1 | 6.7% |
粉红色 |
1 | 6.7% |
粉色 |
1 | 6.7% |
紫色 |
1 | 6.7% |
黑色 |
1 | 6.7% |
Three nearest enrolled references
We compute base-2 JSD for each of the eight categorical distributions, then average the eight distances. A lower value means more similar answer frequencies within this product reference. It does not identify an unknown route by itself.
| Rank | Reference model | Mean 8-cell JSD |
|---|---|---|
| 1 | openai/gpt-4o |
0.700650 |
| 2 | deepseek/deepseek-v3.1-terminus |
0.709763 |
| 3 | openai/gpt-4o-2024-08-06 |
0.711214 |
How many live requests does verification use?
| Profile | Cells × samples | Fresh model requests | Same-data EER |
|---|---|---|---|
| Quick, default complete check | 4 × 5 | 20 | 11.8% |
| Standard | 8 × 5 | 40 | 8.1% |
| Paper budget | 8 × 15 | 120 | 6.7% |
The current-Agent workflow starts with quick. Each answer comes from a fresh Codex or Claude Code process that reuses host-managed login. These requests can consume host allowance or provider balance.
What this fingerprint can and cannot show
- The reference records behavior observed through OpenRouter in the author dataset. Provider wrappers, model updates, decoding, and date can move a distribution.
- The author collection set temperature 1 and disabled reasoning. Current-Agent mode labels sampling as a host default because supported Agent CLIs expose no temperature flag.
- The eight high-gap cells and product thresholds use the same public cohort. Their EER values are calibration results, not an independent production accuracy claim.
- A nearest reference is a similarity result. VerifyLLMAPI returns INCONCLUSIVE when the claimed route lacks an enrolled fingerprint.
Primary sources, license, and reproducibility
- Tomas Bruckner, One Token Is Enough: Fingerprinting and Verifying Large Language Models from Single-Token Output Distributions, arXiv:2607.10252v1.
- Author dataset and derived results, DOI 10.5281/zenodo.21278557, CC BY 4.0.
- Author collection and analysis software, DOI 10.5281/zenodo.21278793, MIT.
- VerifyLLMAPI atlas JSON, built from reference
pamela-openrouter-2026-07-8cell-ref-a, SHA-25651f3f63cbcfb56b151055d2d10e36a9cc2f0644cc68b580784853d68951fa866.
Page evidence SHA-256: f17f8992df172a7e84c8825e02efd77252e1e73ca86391c9460f4d4dd8ebc9d2. This hash covers the model ID, eight cell metrics, nearest references, and strongest cell difference in the generated atlas.