Mastermind
Single-turn Mastermind environment with information-gain / elimination scoring
- Domain
- rl-env
- License
- unknown
- Published
- Sep 2025
Cite
Notes
Only stored in your browser.
Top score 29.1% by Qwen3 Next 80B A3B Instruct - 6 models reporting (1 frontier)
Score history
6Top models
6
Related tools
1Implementations, trainers, datasets and scaffolds linked to this eval.
FAQ
- What is Mastermind?
- Single-turn Mastermind environment with information-gain / elimination scoring
- What is the current top score on Mastermind?
- The top reported score is 29.1% by Qwen3 Next 80B A3B Instruct, across 6 models reporting (1 from frontier labs).
- How can a model improve its Mastermind score?
- Tools linked to Mastermind on Sophon include Mastermind RL Env (Community) - RL environments, datasets, and scaffolds that target this eval.
- What license is Mastermind under?
- Mastermind is available under unknown.