Qwen-3-14B

The 'Qwen-3-14B' dataset/benchmark contains a collection of 300 individual verifiable environments used to evaluate and enhance the reasoning capabilities of Large Language Models through reinforcement learning.

Papers using Qwen-3-14B (3)

From Data to Behavior: Predicting Unintended Model Behaviors Before Training2026Verifiable Environments Are LEGO Bricks: Recursive Composition for Reasoning Generalization2026LLM Compression by Block Removal with Constrained Binary Optimization2026