Archive
Research
-
Completed 12k Run: Early Speech Emergence in a 21.9M OLMo-Hybrid Speech LM
-
A 21.9M hybrid recurrent attention to learn speech on 1.8k steps
-
OLMo Hybrid for Autoregressive Music Token Modeling: A 300M Single-A100 Baseline
-
Janky memory cache over mamba hybrid for music?
-
Transport Geometry as a Testable Hypothesis?
-
Your SAE looks solved. Your model disagrees
-
Building out a mini-compute cluster
-
Review on audio generative systems
-
Notes on speech from signal processing perspective
Essays And Notes
-
Topological Signal In Learned Odor Embeddings
-
Hmm1
-
Consumer llms are coming (here?)
-
A Dirge for the Romantics
-
Neural Fields are cool
-
Latent Diffusion models are Jungian artifacts
-
Language?
-
Synthesis of Culture
-
Hyperindividualism
-
Cybercratic Imperialism
-
A computationalist view of selection
-
On money and fitness
-
On Gans