The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation
Paper • 2609.36484 • Published • 578
Marco Ferrari
marcoferr
·
AI & ML interests
graph neural networks, code generation, deep learning, generative ai
Recent Activity
updated a model 1 day ago
marcoferr/neural-architecture-search published a model 1 day ago
marcoferr/neural-architecture-search liked a model 1 day ago
autotrust/JEV-27B-VLOrganizations
None yet