Theorist calls random-network Lyapunov spectrum paper his biggest breakthrough, co-authored with GPT-6 and Claude
StartupYou · x · 2026-10-09
Theoretical neuroscientist David G. Clark published "Lyapunov spectrum of random neural networks" (arXiv:2610.12426), calling it his most significant pure-theory breakthrough.
- The Lyapunov spectrum — growth rates of perturbations across directions — richly and diffeomorphism-invariantly characterizes chaos
- The paper exactly computes the spectrum for a nonlinear recurrent network with random asymmetric couplings in the N→∞ limit, via a finite-N identity and a cavity-method self-consistent single-site problem
- It establishes that the network's chaos is extensive
- The paper notes the work was done in collaboration with GPT-6 Astra and Claude Opus 5.5 — a notable signal of AI co-authorship in theory research
More from AGI Musings
- VC: Still needing Vercel, Supabase and AWS proves AI has a long way to go — StewartalsopIII · 2026-10-23
- The 'antifragile internet' case: AI agents slash attack costs toward zero — rebeccakaden · 2026-10-23
- Math PhD student quits academia: AI is making math unrecognizable — pickover · 2026-10-23
- Microsoft: China-aligned hackers used ARTEX AI agents to hit 2,250+ M365 tenants — rickasaurus · 2026-10-23
- Andreessen: AI-Native Startups vs Incumbents Show Widest Perception Gap He's Ever Seen — juanbenet · 2026-10-23
- Only 2% of Americans pay for AI, predicted to hit 50% in three years — JOBhakdi · 2026-10-23