A 27M parameter model beating LLMs on reasoning tasks

by SteadySurfdom | Nov 28, 2025 | 1 comments on HN