← Gemini 3.1 Pro · API · High thinking
Evidenceacceptedindependent_lab
AA-Omniscience
Artificial Analysis — AA-Omniscience Index (Gemini 3.1 Pro Preview)
ev_2026_07_gem31pro_aa_omniscience
Result
33
Confidence B
- Tested
- 2026-07-11
- Runs
- 1
Configuration
How it was measured
- Exact model
- gemini-3.1-pro-preview
- Reasoning
- high
- Agent harness
- none
- Tool access
- AA default knowledge probe (not web-search agent)
- Prompting
- AA Omniscience protocol
PRIMARY WIS — AA-Omniscience Index leadership narrative. Gemini 3.1 Pro has repeatedly led Index (~33) among frontiers via lower hallucination vs high-accuracy/high-hall peers (e.g. Sol). Exact live float chart-only on AA.
Caveats
Known limitations
- Index float transcribed from AA article / peer-sheet cross-check; re-scrape chart at freeze.
- Always pair Index with accuracy + hallucination narrative on the sheet.
- normalisedScore is curator judgment for fusion (peer-leading calibration story).
Weighting
Evidence weight factors
The engine weights each evidence row by these factors (0–1) when fusing scores — higher is more trusted.
relevance0.95
quality0.90
recency1.00
health0.85
comparability0.80
reliability0.70
Benchmark
AA-Omniscience
Knowledge reliability benchmark that jointly scores accuracy and hallucination / overconfidence. Rewards calibrated abstention rather than pure guess accuracy.
Benchmark source →