rekaai/reka-edge behavioral LLM fingerprint
This page breaks down the exact reference that VerifyLLMAPI uses for rekaai/reka-edge: 116 normalized answers across eight low-entropy probes, plus its nearest enrolled models by mean Jensen–Shannon divergence.
Fingerprint summary
- Reference answers
- 116
- Samples per cell
- 12–15
- Nearest model
- mistralai/mistral-nemo
- Nearest mean JSD
- 0.854661
The most concentrated cell was letter-random|zh: 七 held 46.7% of its answers and entropy measured 2.597 bits. The broadest cell was word-random|en at 3.807 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 116/120 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 · 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 · 12/15 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. 15 valid reference answers produced 13 normalized values. The mode was dog at 2/15 (13.3%); entropy was 3.640 bits.
| Normalized response | Count | Share |
|---|---|---|
4 |
1 | 6.7% |
7 |
1 | 6.7% |
dog |
2 | 13.3% |
elephant |
2 | 13.3% |
bear |
1 | 6.7% |
cat |
1 | 6.7% |
fox |
1 | 6.7% |
goat |
1 | 6.7% |
horse |
1 | 6.7% |
llama |
1 | 6.7% |
squirrel |
1 | 6.7% |
tiger |
1 | 6.7% |
zebra |
1 | 6.7% |
2. word-random|en
Random word, English. 14 valid reference answers produced 14 normalized values. The mode was book at 1/14 (7.1%); entropy was 3.807 bits.
| Normalized response | Count | Share |
|---|---|---|
book |
1 | 7.1% |
cloud |
1 | 7.1% |
ensure |
1 | 7.1% |
entry |
1 | 7.1% |
fin |
1 | 7.1% |
fossil |
1 | 7.1% |
hand |
1 | 7.1% |
hover |
1 | 7.1% |
lamp |
1 | 7.1% |
male |
1 | 7.1% |
ramp |
1 | 7.1% |
sand |
1 | 7.1% |
school |
1 | 7.1% |
shadow |
1 | 7.1% |
3. animal-random|ru
Random animal, Russian. 15 valid reference answers produced 13 normalized values. The mode was 7 at 3/15 (20.0%); entropy was 3.590 bits.
| Normalized response | Count | Share |
|---|---|---|
3 |
1 | 6.7% |
4 |
1 | 6.7% |
7 |
3 | 20.0% |
8 |
1 | 6.7% |
9 |
1 | 6.7% |
tiger |
1 | 6.7% |
белка |
1 | 6.7% |
газель |
1 | 6.7% |
гусь |
1 | 6.7% |
дмурь |
1 | 6.7% |
лев |
1 | 6.7% |
лис |
1 | 6.7% |
ловушка |
1 | 6.7% |
4. animal-random|ar
Random animal, Arabic. 12 valid reference answers produced 10 normalized values. The mode was الخفر at 2/12 (16.7%); entropy was 3.252 bits.
| Normalized response | Count | Share |
|---|---|---|
الخفر |
2 | 16.7% |
لا |
2 | 16.7% |
الأرجوم |
1 | 8.3% |
الثعلب |
1 | 8.3% |
الخفف |
1 | 8.3% |
الفهاد |
1 | 8.3% |
سيلفي |
1 | 8.3% |
لقم |
1 | 8.3% |
لقمة |
1 | 8.3% |
ميمة |
1 | 8.3% |
5. letter-random|zh
Random character, Chinese. 15 valid reference answers produced 9 normalized values. The mode was 七 at 7/15 (46.7%); entropy was 2.597 bits.
| Normalized response | Count | Share |
|---|---|---|
6 |
1 | 6.7% |
七 |
7 | 46.7% |
i |
1 | 6.7% |
丁 |
1 | 6.7% |
九 |
1 | 6.7% |
十 |
1 | 6.7% |
安 |
1 | 6.7% |
将 |
1 | 6.7% |
陆 |
1 | 6.7% |
6. animal-random|zh
Random animal, Chinese. 15 valid reference answers produced 12 normalized values. The mode was 4 at 2/15 (13.3%); entropy was 3.507 bits.
| Normalized response | Count | Share |
|---|---|---|
1 |
1 | 6.7% |
2 |
1 | 6.7% |
3 |
1 | 6.7% |
4 |
2 | 13.3% |
5 |
1 | 6.7% |
6 |
2 | 13.3% |
7 |
2 | 13.3% |
8 |
1 | 6.7% |
10 |
1 | 6.7% |
99 |
1 | 6.7% |
tiger |
1 | 6.7% |
蛇 |
1 | 6.7% |
7. letter-random|ar
Random letter, Arabic. 15 valid reference answers produced 13 normalized values. The mode was q at 3/15 (20.0%); entropy was 3.590 bits.
| Normalized response | Count | Share |
|---|---|---|
q |
3 | 20.0% |
a |
1 | 6.7% |
c |
1 | 6.7% |
h |
1 | 6.7% |
o |
1 | 6.7% |
t |
1 | 6.7% |
u |
1 | 6.7% |
w |
1 | 6.7% |
x |
1 | 6.7% |
z |
1 | 6.7% |
ا |
1 | 6.7% |
ج |
1 | 6.7% |
ز |
1 | 6.7% |
8. color-random|zh
Random color, Chinese. 15 valid reference answers produced 9 normalized values. The mode was 5 at 4/15 (26.7%); entropy was 2.840 bits.
| Normalized response | Count | Share |
|---|---|---|
1 |
1 | 6.7% |
5 |
4 | 26.7% |
6 |
4 | 26.7% |
7 |
1 | 6.7% |
8 |
1 | 6.7% |
安全 |
1 | 6.7% |
红 |
1 | 6.7% |
被告 |
1 | 6.7% |
转橙 |
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/mistral-nemo |
0.854661 |
| 2 | ai21/jamba-large-1.7 |
0.859098 |
| 3 | google/gemini-2.5-flash |
0.877566 |
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: d8c9471ea5a5ee169d547393f6de630a9eb78630ab4d77abf712ee0277765d5d. This hash covers the model ID, eight cell metrics, nearest references, and strongest cell difference in the generated atlas.