Different Tokens, Different Compute in LLMs @SebastianRaschka
Different Tokens, Different Compute in LLMs  @SebastianRaschka
Uploaded September 2026 | Updated November 2026, 3 hours ago
In Mixture-of-Recursions, a learned router decides which tokens continue through additional passes of the same transformer layers. Some tokens take one pass, others two or three.

An excerpt from my longer video on looped transformers: youtube.com/watch?v=KT4n-z_4QJU
Different Tokens, Different Compute in LLMsL14.3.2.2 ResNet-34 in PyTorch   Code ExampleBuild A Reasoning Model From Scratch 3: The Verifier for Evaluation and RL with Verifiable RewardsL4.1 Tensors in Deep LearningL10.1 Techniques for Reducing OverfittingOpenAI Astra and Recurrent Depth / Looped Transformers13.4.5 Sequential Feature Selection   Code Examples (L13: Feature Selection)L15.7 An RNN Sentiment Classifier in PyTorchBuild A Reasoning Model From Scratch 1: Motivation & Code SetupL8.6 Multinomial Logistic Regression / Softmax RegressionWhere verifiable rewards for reasoning model training come fromHow top-p filters tokens
Sebastian Raschka |

Different Tokens, Different Compute in LLMs

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER