Patrik Wolf
patrikwolf
·
AI & ML interests
Test-time training, preference learning, alignment, theory
Recent Activity
upvoted a paper 5 days ago
Sanity Checks for Sparse Autoencoders: Do SAEs Beat Random Baselines? upvoted a paper 10 days ago
Learning beyond Teacher: Generalized On-Policy Distillation with Reward Extrapolation upvoted a paper 25 days ago
Reinforcement Learning via Self-Distillation Organizations
