Linear-LLM-SCM: Benchmarking LLMs for Coefficient Elicitation in Linear-Gaussian Causal Models
Kanta Yamaoka, Sumantrak Mukherjee, Thomas Gärtner, David Selby, Stefan Konigorski, Eyke Hüllermeier, Viktor Bengs, Sebastian Vollmer · 2026
Read Linear-LLM-SCM: Benchmarking LLMs for Coefficient Elicitation in Linear-Gaussian Causal Models by Kanta Yamaoka, Sumantrak Mukherjee, Thomas Gärtner, David Selby, Stefan Konigorski, Eyke Hüllermeier, Viktor Bengs, Sebastian Vollmer (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-10282.txt.
Description
Language models show promise for quality-style causal relations, but number-based causal thought in continuous domains remains little explored. A 2026 paper led by Kanta Yamaoka, with seven co-authors, offers Linear-LLM-SCM, a plug-and-play test setup for LLMs on linear Gaussian structural causal model setup when a directed acyclic graph is given. The framework splits a DAG into local parent-child groups and asks an LLM to write a regression-style structural equation per node, then pools and compares them against ground-truth parameters.
Tests with seven real-world DAGs with ground truth expose limits of LLMs as number-based causal parameterizers. Across most models, coefficient guesses vary and prove touchy to structural shifts. The framework is open-sourced to spur further work on LLM causal effect elicitation in safety-critical domains such as healthcare.
Catalog record
| CREATOR | Kanta Yamaoka, Sumantrak Mukherjee, Thomas Gärtner, David Selby, Stefan Konigorski, Eyke Hüllermeier, Viktor Bengs, Sebastian Vollmer |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 33 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-10282