Uni-LaDiR: Latent Diffusion Unifies Multimodal Reasoning
One world, many observations. Why we want a shared workspace for reasoning—and why its next state should be a distribution.
Notes & essays
Blog
Notes on reasoning, learning, and building AI systems.
LaDiR: Latent Diffusion Enhances LLMs for Text Reasoning
Why use diffusion to model reasoning? The ideas behind LaDiR: generating thoughts in a learned latent space, training on the model’s own trajectories, and using denoising to control inference-time computation.
No posts match your search. Try another keyword or topic.