Controlled study: LLMs struggle to recover latent sequential structure despite long context
illinois · hf · 2026-10-09
A University of Illinois controlled study using Rock-Paper-Scissors and stochastic n-gram tasks finds LLMs struggle to recover latent sequential structure: longer context doesn't improve strategy identification, correct recognition doesn't ensure faithful simulation, and higher-order dependencies substantially degrade rule recovery—apparent behavioral fidelity can mask incorrect generative mechanisms.
More from Research
- Open differential geometry book by Pinkall & Gross joins ChapterPal with an AI tutor — burkov · 2026-10-09
- RSI-Exam lands in State of AI report: top model Opus 5.5 scores just ~0.53 — cihangxie · 2026-10-09
- KAIST's ME-World tackles multi-agent egocentric world modeling with joint denoising — kaist-ai · 2026-10-09
- TerraVis quantifies world-grounded visual consistency failures in text-to-image models — the-aiml · 2026-10-09
- OneSearch-VL: unified multimodal deep research agent beats Qwen3-VL by 20 points — Hongyu Li · 2026-10-09
- SpaceCast-Bench: best VLM hits 58.0% on predictive spatial reasoning vs 87.2% human — zju · 2026-10-09