2025​

• January 7 - Training a Reasoning Model for Very Cheap with DPO
• May 22 - Training Tiny LLMs to Play Connections Better than Frontier Models

2024​

• January 28 - Understanding StyleAligned
• February 24 - SuperPrompt - Better SDXL prompts in 77M Parameters
• March 11 - Tiny Agents - Training Small LLMs to Use Tools with DPO and Synthetic Data
• November 28 - Curriculum Learning with DPO and Logit Pruning
• December 13 - Reviewing Post-Training Techniques from Recent Open LLMs

2023​

• December 20 - Longer videos with Stable Video Diffusion via YaRN