New paper asks what causal strategies LLMs can learn passively from text
AndrewLampinen · x · 2026-09-29
Andrew Lampinen's team released Passive learning of active causal strategies in agents and language models, engaging the famous octopus thought experiment: how much about causality and experimentation can LLMs learn just by passively imitating text? The paper provides specific claims and empirical evidence. In follow-up discussion, the author adds that at least some kinds of math fall within natural language distributions, and that while math may not convey more meaning, it certainly adds precision.
More from Research
- Triangle Splatting SLAM: Imperial College's ECCV 2026 dense RGB-D SLAM with on-the-fly mesh extraction — rsasaki0109 · 2026-09-30
- Manifold opens early access: robotics eval platform runs thousands of GPU-parallel rollouts in 30 mins — paigeinsf · 2026-09-30
- 1,000 AI agents discover new CRISPR-like system in virus DNA within 24 hours — CurieuxExplorer · 2026-09-30
- Explaining just 5% of token positions retains nearly all audit success across 4.7M explanations — aisilab · 2026-09-30
- NTU's Persistence Forcing hits FID 1.63 on ImageNet 256 by heterogeneous refinement in pixel-space DiTs — NanyangTechnologicalUniversity · 2026-09-30
- IBM's Q&D trains proactive agents to ask better questions, beating a 15x larger model — ibm · 2026-09-30