-
Dario's Big Blog of Compute Hypothesis
My reading of Dario Amodei’s Big Blob of Compute hypothesis: a summary of its seven key factors, followed by the supplied original essay.
-
Understanding Drifting Models Through Distribution Matching
Understanding one-step Drifting Models through score differences, kernel estimation, and distribution matching.
-
What Order Do DeepSeek-OCR 2's Causal Flow Queries Actually Learn?
Attention probes and query-token interventions to investigate the ordering learned by DeepSeek-OCR 2.
-
Distribution Matching VAE (DMVAE): Exploring the Best Latent Space for Visual Generation
Distribution matching for visual tokenization: the motivation, method, and experiments behind DMVAE.