Similar Links:

• Text Compression as a Test for Artificial Intelligence

• Language Modeling Is Compression

• Observational Scaling Laws and the Predictability of Language Model Performance

• U-PaLM: Transcending Scaling Laws with 0.1% Extra Compute

• MMLU: Measuring Massive Multitask Language Understanding

• The Scaling Hypothesis

• Scaling ‘diminishing returns’

• AI Cannibalism Can Be Good

• Towards Better LLM Creative Writing

• Virtual comments: LLM idea

• ‘Tools For Thought’ Failure

• Nenex: A Neural Personal Wiki Idea

• LLM Time Travel Visualization

• First Tragedy, then Parse: History Repeats Itself in the New Era of Large Language Models

• [GPT-5 was a <100× GPT-4 scaleup]

• The Bitter Lesson

• The Hyperfitting Phenomenon: Sharpening and Stabilizing LLMs for Open-Ended Text Generation

• Reconsidering the Past: Optimizing Hidden States in Language Models

• Notes on a New Philosophy of Empirical Science

• Search: GS; Google; site