Recent Posts

• GEPA: Reflective prompt evolution can outperform reinforcement learning
  by che_shr_caton 7/31/2025, 10:57:58 AM with 7 comments
• Subliminal Learning: Language models transmit behavioral traits via hidden
  by che_shr_caton 7/29/2025, 3:33:36 PM with 0 comments
• AlphaGo Moment for Model Architecture Discovery
  by che_shr_caton 7/29/2025, 10:50:21 AM with 0 comments
• Paper FOMO and ICML 2025 Outstanding Papers
  by che_shr_caton 7/28/2025, 8:03:46 PM with 0 comments
• Early Signs of Steganographic Capabilities in Frontier LLMs
  by che_shr_caton 7/8/2025, 1:16:15 PM with 0 comments
• Musicality in Animals
  by che_shr_caton 6/18/2025, 1:26:36 PM with 0 comments
• Text-to-LoRA Enables On-the-Fly Model Adaptation
  by che_shr_caton 6/12/2025, 8:32:10 PM with 0 comments
• Quantum computing and artificial intelligence: status and perspectives
  by che_shr_caton 6/4/2025, 1:00:35 PM with 0 comments
• The Most Misunderstood Feature of the Sound
  by che_shr_caton 6/2/2025, 12:57:19 PM with 0 comments
• Darwin Gödel Machine
  by che_shr_caton 6/1/2025, 4:02:19 PM with 0 comments
• Are Deeper LLMs Smarter, or Just Longer?
  by che_shr_caton 5/27/2025, 12:40:58 PM with 0 comments
• Muon Optimizer Accelerates Grokking
  by che_shr_caton 4/28/2025, 12:40:27 PM with 0 comments
• ThoughtTerminator
  by che_shr_caton 4/24/2025, 9:43:54 PM with 0 comments
• Chain of Continuous Thought (Coconut)
  by che_shr_caton 4/22/2025, 5:03:49 PM with 0 comments
• Intuitive Physics Emergence in V-JEPA
  by che_shr_caton 4/4/2025, 4:53:22 PM with 0 comments
• Sound physics And basics of sound perception
  by che_shr_caton 4/3/2025, 1:55:20 PM with 0 comments
• BLT: Byte Latent Transformer
  by che_shr_caton 12/26/2024, 12:18:49 AM with 0 comments
• A Single 'Super Weight' Can Break Your Billion-Parameter Model
  by che_shr_caton 11/29/2024, 3:00:34 PM with 0 comments
• Jax Things to Watch for in 2025
  by che_shr_caton 11/26/2024, 2:37:28 PM with 0 comments
• Diffusion models are evolutionary algorithms
  by che_shr_caton 11/9/2024, 10:51:27 PM with 5 comments
• Make Softmax Great Again
  by che_shr_caton 11/6/2024, 11:19:07 PM with 0 comments
• Deep Learning Frameworks: The Fourth Pillar of Deep Learning Revolution
  by che_shr_caton 11/5/2024, 6:01:02 PM with 0 comments
• TextGrad: Automatic "Differentiation" via Text
  by che_shr_caton 6/26/2024, 8:40:18 PM with 0 comments
• Superconducting Supercomputers
  by che_shr_caton 6/24/2024, 10:30:52 PM with 0 comments
• Decoder-decoder architecture is coming
  by che_shr_caton 6/2/2024, 1:26:59 PM with 0 comments
• Chronos: Using Pretrained LLMs for Probabilistic Time Series Forecasting
  by che_shr_caton 4/28/2024, 6:52:48 PM with 0 comments
• Big Post About Big Context
  by che_shr_caton 2/29/2024, 5:59:16 PM with 3 comments
• Neural Network Diffusion
  by che_shr_caton 2/26/2024, 12:47:41 PM with 0 comments
• Thermodynamic AI is getting hotter
  by che_shr_caton 2/8/2024, 10:53:10 AM with 2 comments
• Training LLMs with AMD GPUs on Frontier Supercomputer
  by che_shr_caton 1/16/2024, 1:08:48 PM with 0 comments
• Beyond Chinchilla-Optimal Accounting for Inference in Language Model Scaling Law
  by che_shr_caton 1/8/2024, 10:03:11 PM with 0 comments
• Project CETI
  by che_shr_caton 12/17/2023, 7:36:19 PM with 0 comments
• GonzoML on Mamba and S6 (+previous post on S4)
  by che_shr_caton 12/13/2023, 1:56:53 PM with 0 comments
• Conway's Game of Life Is Omniperiodic
  by che_shr_caton 12/9/2023, 11:29:12 AM with 1 comment
• GonzoML on Gemini
  by che_shr_caton 12/7/2023, 5:14:15 PM with 0 comments
• Matryoshka Representation Learning
  by che_shr_caton 11/3/2023, 10:28:31 PM with 0 comments
• Mindstorms in Natural Language-Based Societies of Mind
  by che_shr_caton 10/29/2023, 12:43:34 PM with 0 comments
• The convolution empire strikes back
  by che_shr_caton 10/27/2023, 7:39:31 PM with 6 comments
• Sparse Universal Transformer
  by che_shr_caton 10/23/2023, 8:16:33 PM with 0 comments
• MemWalker: An alternative way for working with long documents using transformers
  by che_shr_caton 10/17/2023, 12:56:09 PM with 0 comments
• "Building Machines That Learn and Think Like People", 7 Years Later
  by che_shr_caton 10/13/2023, 9:39:58 PM with 8 comments
• Chain-of-Thought → Tree-of-Thought
  by che_shr_caton 10/10/2023, 5:42:25 PM with 0 comments
• Mortal Computers
  by che_shr_caton 10/9/2023, 1:25:09 PM with 1 comment
• Levanter – Legible, Scalable, Reproducible Foundation Models with Jax
  by che_shr_caton 6/20/2023, 1:40:44 PM with 0 comments
• LM-3 –- resurrecting the MIT CADR
  by che_shr_caton 5/12/2023, 3:57:13 PM with 0 comments
• The Annotated Diffusion Model
  by che_shr_caton 6/13/2022, 10:27:25 AM with 0 comments