Source drift radar
Source changes become review signals.
Schema movement, row disappearance, column rename/parser changes, and unusual score shifts are grouped by source.
| Source | Severity | Captured | Parser | Status | Reasons |
|---|---|---|---|---|---|
| LiveBench | ok | 2026-06-24 | 0.5.0 | verified | No drift warning |
| MTEB | ok | 2026-06-24 | 0.1.0 | verified | No drift warning |
| OpenCompass | ok | 2026-06-24 | 0.1.0 | verified | No drift warning |
| Provider official evals | ok | 2026-06-24 | unrecorded | verified | No drift warning |
| Scale Labs | ok | 2026-06-24 | 0.1.0 | verified | No drift warning |
| Arena | warning | 2026-06-24 | 0.5.0 | verified | source snapshot changed: Source snapshot changed · schema or parser changed: Parser output changed |
| Artificial Analysis | warning | 2026-06-24 | 0.5.0 | verified | source snapshot changed: Source snapshot changed · schema or parser changed: Parser output changed |
| LLMBase | warning | 2026-06-24 | 0.2.0 | relay | source snapshot changed: Source snapshot changed |
| Terminal-Bench | warning | 2026-06-24 | 0.2.0 | verified | source snapshot changed: Source snapshot changed |
| Vals AI | warning | 2026-06-24 | 0.1.0 | verified | source snapshot changed: Source snapshot changed · schema or parser changed: Parser output changed |