Prismatic Synthesis: Gradient-based Data Diversification Boosts Generalization in LLM Reasoning

NeurIPSSpotlight2025

Authors
Jaehun Jung, Seungju Han, Ximing Lu, Skyler Hallinan, David Acuna, Shrimai Prabhumoye, Mostofa Patwary, Mohammad Shoeybi, Bryan Catanzaro, Yejin Choi
Venue
NeurIPS 2025
Track
Spotlight

TL;DR

Data diversity is crucial for training a strong language model…

Opening excerpt from the authors’ abstract. source

Read the paper

Topics

generalization language model reasoning llm

Related NeurIPS papers

← All NeurIPS 2025 Spotlight papers · Browse the whole archive