PA1-OPD: One-Step Prefix-Augmented On-Policy Distillation for Language Model Pretraining

arXiv:2610.19601 · cs.CL · Submitted 2026-10-14 · Read on arXiv

cs.CL

Submitted: 2026-10-14

Updated: 2026-10-14

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Distillation — 1× in this paper · explained in 2 episodes
• On-Policy Distillation — 1× in this paper · explained in 3 episodes

Sources

• The Pile: An 800GB Dataset of Diverse Text for Language Modeling
• Gemma 2: Improving Open Language Models at a Practical Size
• Distilling the Knowledge in a Neural Network
• SelecTKD: Selective Token-Weighted Knowledge Distillation for LLMs
• Compact Language Models via Pruning and Knowledge Distillation
• Compressive Transformers for Long-Range Sequence Modelling
• DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter
• 2 OLMo 2 Furious
• Efficient Pretraining Length Scaling
• Cache-Aware Joint Router Adaptation for Memory-Efficient MoE Inference
• Prefix-Guided On-Policy Distillation: Mining Golden Trajectories from Rollouts

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering