upstage/solar-pro-3 behavioral LLM fingerprint
This page breaks down the exact reference that VerifyLLMAPI uses for upstage/solar-pro-3: 117 normalized answers across eight low-entropy probes, plus its nearest enrolled models by mean Jensen–Shannon divergence.
Fingerprint summary
- Reference answers
- 117
- Samples per cell
- 14–15
- Nearest model
- mistralai/voxtral-small-24b-2507
- Nearest mean JSD
- 0.617707
The most concentrated cell was animal-random|en: elephant held 40.0% of its answers and entropy measured 1.839 bits. The broadest cell was letter-random|zh at 3.807 bits.
Against the closest enrolled model, letter-random|zh contributed the largest cell distance at JSD 0.855361. 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 117/120 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 · 14/15 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 · 14/15 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 · 14/15 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. 15 valid reference answers produced 5 normalized values. The mode was elephant at 6/15 (40.0%); entropy was 1.839 bits.
| Normalized response | Count | Share |
|---|---|---|
elephant |
6 | 40.0% |
lion |
6 | 40.0% |
dog |
1 | 6.7% |
giraffe |
1 | 6.7% |
panda |
1 | 6.7% |
2. word-random|en
Random word, English. 15 valid reference answers produced 8 normalized values. The mode was serendipity at 4/15 (26.7%); entropy was 2.740 bits.
| Normalized response | Count | Share |
|---|---|---|
serendipity |
4 | 26.7% |
apple |
3 | 20.0% |
ocean |
3 | 20.0% |
glimmer |
1 | 6.7% |
purple |
1 | 6.7% |
rainbow |
1 | 6.7% |
sunset |
1 | 6.7% |
回答は「oeis」 |
1 | 6.7% |
3. animal-random|ru
Random animal, Russian. 14 valid reference answers produced 12 normalized values. The mode was собака at 3/14 (21.4%); entropy was 3.468 bits.
| Normalized response | Count | Share |
|---|---|---|
собака |
3 | 21.4% |
elephant |
1 | 7.1% |
giraffe |
1 | 7.1% |
gorilla |
1 | 7.1% |
lion |
1 | 7.1% |
panda |
1 | 7.1% |
rabbit |
1 | 7.1% |
животное |
1 | 7.1% |
коала |
1 | 7.1% |
кошка |
1 | 7.1% |
леопард |
1 | 7.1% |
слон |
1 | 7.1% |
4. animal-random|ar
Random animal, Arabic. 14 valid reference answers produced 10 normalized values. The mode was lion at 5/14 (35.7%); entropy was 2.978 bits.
| Normalized response | Count | Share |
|---|---|---|
1 |
1 | 7.1% |
lion |
5 | 35.7% |
cat |
1 | 7.1% |
elephant |
1 | 7.1% |
الديناصور< |
1 | 7.1% |
الغزال |
1 | 7.1% |
النمر |
1 | 7.1% |
بِلْبل</b> |
1 | 7.1% |
ق众多 |
1 | 7.1% |
نمر |
1 | 7.1% |
5. letter-random|zh
Random character, Chinese. 14 valid reference answers produced 14 normalized values. The mode was 二 at 1/14 (7.1%); entropy was 3.807 bits.
| Normalized response | Count | Share |
|---|---|---|
二 |
1 | 7.1% |
字 |
1 | 7.1% |
日 |
1 | 7.1% |
明 |
1 | 7.1% |
朔 |
1 | 7.1% |
歪 |
1 | 7.1% |
流 |
1 | 7.1% |
漢字 |
1 | 7.1% |
白 |
1 | 7.1% |
阳 |
1 | 7.1% |
随机汉字 |
1 | 7.1% |
风 |
1 | 7.1% |
魏 |
1 | 7.1% |
龙 |
1 | 7.1% |
6. animal-random|zh
Random animal, Chinese. 15 valid reference answers produced 12 normalized values. The mode was elephant at 3/15 (20.0%); entropy was 3.457 bits.
| Normalized response | Count | Share |
|---|---|---|
elephant |
3 | 20.0% |
猫 |
2 | 13.3% |
blue |
1 | 6.7% |
frog |
1 | 6.7% |
lion |
1 | 6.7% |
企鹅 |
1 | 6.7% |
大象 |
1 | 6.7% |
熊猫 |
1 | 6.7% |
狐狸 |
1 | 6.7% |
狮子 |
1 | 6.7% |
猴子 |
1 | 6.7% |
鳄鱼 |
1 | 6.7% |
7. letter-random|ar
Random letter, Arabic. 15 valid reference answers produced 11 normalized values. The mode was ب at 3/15 (20.0%); entropy was 3.323 bits.
| Normalized response | Count | Share |
|---|---|---|
ب |
3 | 20.0% |
a |
2 | 13.3% |
أ |
2 | 13.3% |
e |
1 | 6.7% |
o |
1 | 6.7% |
q |
1 | 6.7% |
x |
1 | 6.7% |
ا |
1 | 6.7% |
د |
1 | 6.7% |
م |
1 | 6.7% |
ژ |
1 | 6.7% |
8. color-random|zh
Random color, Chinese. 15 valid reference answers produced 6 normalized values. The mode was 红色 at 8/15 (53.3%); entropy was 1.990 bits.
| Normalized response | Count | Share |
|---|---|---|
红色 |
8 | 53.3% |
red |
3 | 20.0% |
1f53ff |
1 | 6.7% |
紅 |
1 | 6.7% |
绿色 |
1 | 6.7% |
蓝色 |
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 | mistralai/voxtral-small-24b-2507 |
0.617707 |
| 2 | amazon/nova-2-lite-v1 |
0.665028 |
| 3 | amazon/nova-premier-v1 |
0.673472 |
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: 02c78f1fba1084eb011f8775af31486cb7b48cc76c9bce9f1f1364c635430472. This hash covers the model ID, eight cell metrics, nearest references, and strongest cell difference in the generated atlas.