Categories

Browse by Category

Click on any category to search for posts in that category.

All Posts by Category

AGI

Deep Equilibrium Models and the Implicit Function Theorem

The Hierarchical Reasoning Model

Understanding Recurrence in Modern Models

ARC-AGI 3 2025 July Demo

Coding World Models with LLMs

ARC-AGI 3 - Gaming with Reason

Normalising a Short AGI Horizon

Understanding Human Thought: Perspectives from Cognitive Science and AI

Test-Time Compute Matters: From LLMs to Search

LLM Readability as a Tool

Matching Robots

Why Autonomous Software?

Maintaining Old Work

AGI as the Best Gamer

How to Evaluate Reasoning Capabilities

What is Reasoning?

Learning Efficiently

A Basic Framework for General Intelligence

GPUs are Awesome

Another Great Year for AI, 2024 -> 2025

AI-Generated

Understanding Human Thought: Perspectives from Cognitive Science and AI

Competition

ARC-AGI 3 2025 July Demo

ARC-AGI 3 - Gaming with Reason

Experiment

I Sense a Contradiction

Stacking at Length

Knowledge

Understanding Recurrence in Modern Models

Normalising a Short AGI Horizon

Test-Time Compute Matters: From LLMs to Search

How to Run Cheap LLM Experiments

Why Autonomous Software?

Four Methods to Visualise Programs

Grammatically Correct

LLMs

How to Run Cheap LLM Experiments

The DeepSeek-R1 Training Pipeline

Why we can't talk to dogs, yet

LLM Readability as a Tool

AGI as the Best Gamer

Saturated Benchmarks Are Not Solved Benchmarks

I Sense a Contradiction

Autoregressive Errors are Independent?

LLM Architecture Scales

Stacking at Length

Grammatically Correct

o3 is Coming

LLMs as a Repository of Vector Programs

News

AGI as the Best Gamer

o3 is Coming

Productivity

Visualising Code Diffs

Hello, Obsidian!

Think it Faster in AI

How to Run Cheap LLM Experiments

Projects

Why we can't talk to dogs, yet

RL

Matching Robots

Rant

Maintaining Old Work

Saturated Benchmarks Are Not Solved Benchmarks

Review

Deep Equilibrium Models and the Implicit Function Theorem

The Hierarchical Reasoning Model

The DeepSeek-R1 Training Pipeline

Tutorial

Reconstructing Pixel Blocks

Coding World Models with LLMs