← Search

Deyang Kong

2 accepted papers

2026

Rethinking the Sampling Criteria in Reinforcement Learning for LLM Reasoning: A Competence-Difficulty Alignment Perspective

AAAI 2026technical

The low sampling efficiency during the rollout phase poses a significant challenge to scaling reinforcement learning for large language model reasoning. Existing methods attempt to improve efficiency by scheduling problems based on problem difficulties. However, these approaches suffer from unstabl

Cited by 0SourcePDFScholar
2025

SampleMix: A Sample-wise Pre-training Data Mixing Strategy by Coordinating Data Quality and Diversity

EMNLP 2025

Existing pretraining data mixing methods for large language models (LLMs) typically follow a domain-wise methodology, a top-down process that first determines domain weights and then performs uniform data sampling across each domain. However, these approaches neglect significant inter-domain overlap

Cited by 0SourcePDFScholar