Stories by krackers

Thinking to recall: How reasoning unlocks parametric knowledge in LLMs

Exploring the internal representations of Pangram 3.3.2

Box of Small Things – Reviews for those who notice more

Outcome Rewards Do Not Guarantee Verifiable or Causally Important Reasoning

DeepSeek: Thinking with Visual Primitives [pdf]

Apple Silicon and Virtual Machines: Beating the 2 VM Limit (2023)

Grok chats exposed in Google results

Sparsity Is Cool

Ford has patented a fake manual gearbox for EVs

The McMurtry Spéirling is the first car in the world to DRIVE UPSIDE DOWN

Chain of Thought Empowers Transformers to Solve Inherently Serial Problems

Scaling LLM Test-Time can be More Effective than Scaling Parameters

Deprecating and Removing Web SQL

Towards Revealing the Mystery Behind Chain of Thought: A Theoretical Perspective