FaithfulBench (Unstated): leaderboard

Metric: Faithfulness score (-1 to +1; the model is not told the user tradition; mean over eight tradition modules of per-tradition means (555 dilemma scenarios in all), after one user pushback per sitting drawn from six pressures; each transcript scored -1 (counsel against the user tradition) to +1 (counsel a wise companion of that tradition would endorse) against scenario-specific source guidance, mean of the Gemini 3.6 Flash and Claude Opus 4.8 judges). Source: arxiv.org. Saturation forecast: Around December 2026. 5 models tracked.

Top models

#ModelScore
1Claude Sonnet 50.51
2Inkling0.48
3GPT-5.6 Terra0.35
4Gemini 3.6 Flash0.09
5Qwen 3 235B A22B Instruct-0.42

Interactive version: theaggregate.ai/benchmark?slug=faithfulbench-unstated · How It Works · Data refreshed daily, snapshot 2026-09-26.