Egor Bogomolov
2 papers · Latest:
Machine Learning
Step Rejection Fine-Tuning: A Practical Distillation Recipe
Step Rejection Fine-Tuning (SRFT) improves LLM agent training by leveraging partially correct, unresolved trajectories, outperforming standard RFT.
2605.10674
Natural Language ProcessingFrom Where Words Come: Efficient Regularization of Code Tokenizers Through Source Attribution
This paper introduces Source-Attributed BPE (SA-BPE) to regularize code tokenizers, reducing under-trained tokens caused by data imbalance.
2604.14053
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.