ObviousBench A
AI benchmark · Knowledge
What is ObviousBench?
ObviousBench measures whether language models avoid simple, high-visibility mistakes — the kind of "obvious" errors users notice immediately. It spans eight categories of product-visible failure: arithmetic, character counting, spelling transforms, ordering, negation, format compliance, word counting, and constraint awareness. The headline metric is answer pass³ (all three sampled attempts must be answer-correct), emphasizing reliability across repeated attempts over single-shot accuracy. The public v0.2 release scores 293 model configurations across 144 questions; a held-out private set is not released.
Tier A
Widely used: a lab reports it, or 50 or more models are scored on it.
How much stands behind a tier: usage, who backs it, a confirmed line from its own site or paper, and a clear read on X. High with three of the four, medium with two, low otherwise.
- 238 models scored
- model counts only from community leaderboards
- in curated RL directories (1 pt)
How we got to tier A
- Adoption: 238 models scored; model counts only from community leaderboards. Tier A.
- On X: unclear, none buzz No move.
- In curated RL directories: 1 point, a floor of none.
- Not yet confirmed by a line from its own site or paper.
A clear view on X moves a tier one step up or down; most entries don't move. How tiers work.
Not on the map yet: no line from its own site or paper places it in a domain.
Next: browse all benchmarks. Entry from the RL Research daily scrape of public sources, 2026-10-11 snapshot.