The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation
Paper • 2609.36484 • Published • 45
Grzegorz Kozlowski
grkozlowski
·
AI & ML interests
Into NLP, diffusion models and AI agents.
Recent Activity
upvoted a paper about 9 hours ago
VoxMem: Benchmarking Multimodal Memory in Large Audio Language Models upvoted a paper about 9 hours ago
MaLiang-Harness: A Programmable Path to Image and Video GenerationOrganizations
None yet