Stanford's CausalSmith automates causal inference research with Lean-verified proofs
StanfordAILab · x · 2026-10-15
Stanford AI Lab researchers Jiyuan Tan and Vasilis Syrgkanis released CausalSmith, a formally grounded, self-improving agentic framework that automates theoretical research in causal inference.
- Motivation: LLMs have broken open problems in pure math (OpenAI claims a Lean-verified solution to Navier–Stokes); the team tests whether this extends to applied-science theory.
- Why causal inference: results are precise theorems like pure math, but each theorem depends on modeling real-world assumptions — a harder testbed than pure math.
- Method: the agentic pipeline verifies proofs with Lean and adds an audit step ensuring formal statements match the paper's claims, producing results useful to both theorists and practitioners.
More from coding & agent
- When Should Coding Models Switch from NTP to RL? NTP Overfits Fast, RL Scales — brianryhuang · 2026-10-15
- Starlink's Grok Voice 3.0 agent handles 20,000 daily calls, 70% with no human — elonmusk · 2026-10-15
- Cohere Labs: more coding agents isn't a scaling strategy — when to delegate work — pvncher · 2026-10-15
- Codex tip: ask it to find a specific past conversation in a new chat — petergyang · 2026-10-15
- Designer uses Claude to aggregate years of X, Medium, Dribbble work into one gallery — tomjohndesign · 2026-10-15
- Crimsonland 0.17 adds custom quests you can build by describing ideas to an agent — banteg · 2026-10-15