ModelsLeaderboardEvalsTrainRentalsAPI Docs

Local Reasoning Mini

Official

A lightweight 10-question reasoning sanity check for local models. Tests basic math, logic, and instruction following.

Category: ReasoningRunner: CustomVersion: v1.0.0Submitted by: Lottolabs

Eval Details

Scoring
Exact Match
Aggregation
Mean
Direction
Lower is better
Tasks
2 tasks

Default Run Config

TopP: 1Temperature: 0
TaskDatasetWeightShotsMax Tokens
basic_math
5 inline items1Default16
basic_logic
5 inline items1Default8

Leaderboard— best run per model

No approved results yet. Submit a run via the API.