Philipp Kraehenbuehl

Entropy-preserving reinforcement learning

Triangle Multiplication is All You Need for Biomolecular Structure Representations

Interactive Post-Training for Vision-Language-Action Models

Compressed Map Priors for 3D Perception

PerceptionLM: Open-Access Data and Models for Detailed Visual Understanding

Cut Your Losses in Large-Vocabulary Language Models

Robust Autonomy Emerges from Self-Play

Does Spatial Cognition Emerge in Frontier Models?

Image and Video Tokenization with Binary Spherical Quantization

Language-Image Models with 3D Understanding

Distilling Structural Representations into Protein Sequence Models

3D Perception with Differentiable Map Priors