inception/mercury-2 behavioral LLM fingerprint
This page breaks down the exact reference that VerifyLLMAPI uses for inception/mercury-2: 112 normalized answers across eight low-entropy probes, plus its nearest enrolled models by mean Jensen–Shannon divergence.
Fingerprint summary
- Reference answers
- 112
- Samples per cell
- 11–15
- Nearest model
- openai/gpt-4o-2024-08-06
- Nearest mean JSD
- 0.492926
The most concentrated cell was color-random|zh: 蓝色 held 100.0% of its answers and entropy measured 0.000 bits. The broadest cell was animal-random|ru at 3.418 bits.
Against the closest enrolled model, letter-random|zh 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 112/120 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 · 15/15 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 · 15/15 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 · 12/15 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 · 11/15 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 · 14/15 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. 15 valid reference answers produced 5 normalized values. The mode was elephant at 11/15 (73.3%); entropy was 1.370 bits.
| Normalized response | Count | Share |
|---|---|---|
elephant |
11 | 73.3% |
cat |
1 | 6.7% |
dog |
1 | 6.7% |
elephantassistant |
1 | 6.7% |
otter |
1 | 6.7% |
2. word-random|en
Random word, English. 15 valid reference answers produced 3 normalized values. The mode was serendipity at 13/15 (86.7%); entropy was 0.700 bits.
| Normalized response | Count | Share |
|---|---|---|
serendipity |
13 | 86.7% |
apple |
1 | 6.7% |
xassistant |
1 | 6.7% |
3. animal-random|ru
Random animal, Russian. 12 valid reference answers produced 11 normalized values. The mode was тигр at 2/12 (16.7%); entropy was 3.418 bits.
| Normalized response | Count | Share |
|---|---|---|
тигр |
2 | 16.7% |
жираф |
1 | 8.3% |
жираф<|inc |
1 | 8.3% |
коала |
1 | 8.3% |
конёк |
1 | 8.3% |
корова |
1 | 8.3% |
кошка<|inc |
1 | 8.3% |
кролик |
1 | 8.3% |
лемур |
1 | 8.3% |
слон |
1 | 8.3% |
томат |
1 | 8.3% |
4. animal-random|ar
Random animal, Arabic. 11 valid reference answers produced 4 normalized values. The mode was نمر at 5/11 (45.5%); entropy was 1.677 bits.
| Normalized response | Count | Share |
|---|---|---|
نمر |
5 | 45.5% |
قطة |
4 | 36.4% |
بطة |
1 | 9.1% |
زرافة |
1 | 9.1% |
5. letter-random|zh
Random character, Chinese. 15 valid reference answers produced 9 normalized values. The mode was 字 at 5/15 (33.3%); entropy was 2.866 bits.
| Normalized response | Count | Share |
|---|---|---|
字 |
5 | 33.3% |
林 |
2 | 13.3% |
森 |
2 | 13.3% |
木 |
1 | 6.7% |
爱 |
1 | 6.7% |
福 |
1 | 6.7% |
笑 |
1 | 6.7% |
螺 |
1 | 6.7% |
黏 |
1 | 6.7% |
6. animal-random|zh
Random animal, Chinese. 15 valid reference answers produced 8 normalized values. The mode was 猫 at 6/15 (40.0%); entropy was 2.606 bits.
| Normalized response | Count | Share |
|---|---|---|
猫 |
6 | 40.0% |
狗 |
2 | 13.3% |
猴子 |
2 | 13.3% |
panda |
1 | 6.7% |
企鹅 |
1 | 6.7% |
老虎 |
1 | 6.7% |
虎 |
1 | 6.7% |
豹子 |
1 | 6.7% |
7. letter-random|ar
Random letter, Arabic. 15 valid reference answers produced 7 normalized values. The mode was ب at 6/15 (40.0%); entropy was 2.473 bits.
| Normalized response | Count | Share |
|---|---|---|
ب |
6 | 40.0% |
غ |
2 | 13.3% |
ق |
2 | 13.3% |
ك |
2 | 13.3% |
b |
1 | 6.7% |
س |
1 | 6.7% |
ن |
1 | 6.7% |
8. color-random|zh
Random color, Chinese. 14 valid reference answers produced 1 normalized values. The mode was 蓝色 at 14/14 (100.0%); entropy was 0.000 bits.
| Normalized response | Count | Share |
|---|---|---|
蓝色 |
14 | 100.0% |
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-2024-08-06 |
0.492926 |
| 2 | openai/gpt-4.1-mini |
0.527992 |
| 3 | google/gemma-3n-e4b-it |
0.528795 |
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: dbe5a1ec4846d7177eb1cf534674a8956c83e2dc1f42307f413ad90438fb36a8. This hash covers the model ID, eight cell metrics, nearest references, and strongest cell difference in the generated atlas.