hy4preview · Model identity and integrity report
Last updated：Sep 4, 2026, 5:04 AM · Data window: past 90 days
Measured results (90 days)
5 runs · 3 hosts · 5 anomalies
Family-fingerprint consistency
anthropic 3 ███████████████░░░░░░░░░░ 60% (suspicious) openai 1 █████░░░░░░░░░░░░░░░░░░░░ 20% (suspicious) google 1 █████░░░░░░░░░░░░░░░░░░░░ 20% (suspicious)
Observed V3 sub-model mapping
→ abstain 4 (80%) — low confidence → unknown 1 (20%) ⚠ mismatch
Capability score distribution (auxiliary)
0-49 ███████████████░░░░░░░░░░ 3 50-69 ██████████░░░░░░░░░░░░░░░ 2 70-89 ░░░░░░░░░░░░░░░░░░░░░░░░░ 0 90-100 ░░░░░░░░░░░░░░░░░░░░░░░░░ 0
Host integrity ranking
| Host | Runs | Average score | Family match | Most common V3 match |
|---|---|---|---|---|
| https://api.a6api.com/v1 | 3 | 17 | 0% | — |
| https://a6api.com/v1 | 1 | 51 | 0% | — |
| http://bore.pub:28138/v1 | 1 | 5 | 0% | — |
Recent anomalies
https://api.a6api.com/v1Sep 4, 2026, 4:49 AM
- Self-claimed model
- hy4-preview
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://a6api.com/v1Sep 3, 2026, 6:01 PM
- Self-claimed model
- hy4-preview
- V2 predicted family
- openai
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://api.a6api.com/v1Sep 3, 2026, 4:23 PM
- Self-claimed model
- hy4-preview
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
https://api.a6api.com/v1Sep 3, 2026, 7:44 AM
- Self-claimed model
- hy4-preview
- V2 predicted family
- anthropic
- V3 best match
- —
- Verdict
- V3 low confidence · Abstain
http://bore.pub:28138/v1Aug 31, 2026, 11:24 AM
- Self-claimed model
- am/hy4-preview
- V2 predicted family
- V3 best match
- —
- Verdict
- Family mismatch
Methodology
BazaarLink Probe combines a Bayesian V2 family classifier across nine fingerprint categories with V3 sub-model references to verify endpoints claiming hy4preview. Low family consistency, a different V3 match, or divergence from selfClaim is flagged as anomalous. BazaarLink Probe home.