AI Paper Atlas

Research map / Probabilistic and scientific modeling / Causal discovery

AgentCBO: Causal Belief Routing for Bayesian Optimization under Unknown Graphs

Qiyu Wei, Xin Zhang, Richard Allmendinger, Mauricio A Álvarez

NeurIPS 2026 · Probabilistic and scientific modeling › Causal discovery · approach: Constraint-based structure learning with latents · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes AgentCBO, a causal Bayesian optimization method for unknown causal graphs that routes sampling toward structural uncertainty relevant to the optimizer rather than full graph recovery.

Problem

Causal Bayesian optimization usually assumes a known causal graph; existing relaxations couple optimization with structure learning but waste sample budget on structural uncertainty irrelevant to the optimizer.

Approach

Introduces routing-equivalence, shifting the objective from full causal-graph recovery to identifying the parts of the causal structure that matter for routing optimization decisions, with a belief-based routing mechanism for BO under unknown graphs.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add LLM prior (Language-Prior Agent) instead of PC-only discovery (ablation)AgentCBO · Rosenbrockbest objective Y*: -89.80 -> -90.98Improved
add LLM prior instead of PC-only discovery (ablation)AgentCBO · CompleteGraphbest objective Y*: -2.13 -> -2.50Improved
add LLM prior instead of PC-only discovery (ablation)AgentCBO · PISHCATGraphbest objective Y*: +0.28 -> +0.30No clear effect
add calibration and repair agents on top of the prior (ablation)AgentCBO · CompleteGraphbest objective Y*: -2.50 -> -3.57Improved
add calibration and repair agents on top of the prior (ablation)AgentCBO · ten-benchmark suitebest objective Y*: p=0.84 (Calib), p=0.83 (Full)No clear effect
anonymize or shuffle variable names given to the LLM (ablation)Language-Prior Agent · ten-benchmark suitebest objective Y*Mixed
bypass the deterministic commit gate under a corrupted LLM update (ablation)AgentCBO · PSAGraph stress testendpoint degradation: 0 -> +0.1387Hurt
restrict exploration sets to fixed {B},{D},{B,D} family vs sets containing E (ablation)AgentCBO ES family · CompleteGraphlow-objective hits: 14/20 vs 0/20Improved
use random pair-induced ES family instead of fixed BD family (ablation)AgentCBO ES family · CompleteGraphlow-objective hits: 14/20 -> 4/20Hurt
delete one edge on a manipulative-to-Y path (ablation)AgentCBO with ground-truth graph · Rosenbrockbest objective Y*: -90.97 -> -44.78Hurt
use LLM prior instead of PC-only when coverage rescue is disabled (ablation)AgentCBO · ALARM 37 nodesnormalized endpoint: 1.001 -> 0.23Improved
enable default coverage rescue (ablation)AgentCBO PC-only vs prior · ALARM 37 nodesnormalized endpointNo clear effect

Experimental setup

Builds on: causal Bayesian optimization

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy