Benchmark
Reasoning Gym
A benchmark suite for evaluating the reasoning ability of language models, including tasks such as logic and games.
Timeline (last 90 days)
Nov 2
On the Reasoning Gym benchmark, using the Gemma-3-4B model, CRoS achieves 46.8% accuracy on five unseen logic and game datasets, surpassing the strongest baseline (43.8%).
Unverified50%