Qwen2.5-Math models achieve strong math reasoning by applying self-improvement across pre-training data generation, iterative reward-model-guided SFT and RL, and inference-time sampling.
- Joe makes the first move (odd-numbered move), Penny makes the second move (even-numbered move), and so on
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvement
Qwen2.5-Math models achieve strong math reasoning by applying self-improvement across pre-training data generation, iterative reward-model-guided SFT and RL, and inference-time sampling.