Reasoning slowdowns are transient chaos: paper shows models have fractal basins of attraction
wgilpin0 · x · 2026-09-09
A new arXiv paper (2609.04963) by Jeffrey Lai, Anthony Bao, John Quinn and William Gilpin shows reasoning models exhibit transient chaos — a physical consequence of the computational complexity of hard tasks. Across Sudoku, mazes, visual puzzles and math logic, leading models behave as dynamical systems with fractal basins, fractality increasing with difficulty. Chaos emerges as reasoning gets trapped near saddle points corresponding to nearly-correct solutions, making slowdowns an inevitable cost of problem hardness.
More from Models
- MissionReport benchmark: only 3 of 540 AI-generated government reports pass all criteria — Exp_Mark · 2026-10-27
- Local beats cloud: Qwen 3.8 FN handles private student data as well as frontier models — dr_lm · 2026-10-27
- Aidan Gomez: model's CoT said no tool calls right before making them — aidangomez · 2026-10-27
- Nat Lambert: Cohere's model is narrow RL training on internal infra ill-suited for open weights — natolambert · 2026-10-27
- OpenAI teases DOTS is coming to Europe 'soon' — mark_k · 2026-10-27
- Clifford Pickover challenges AI to decipher a message he encoded in Norse runes — pickover · 2026-10-27