Causal Parrots: Large Language Models May Talk Causality But Are Not Causal
Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel pith:2DAFY5ASrecord.jsonopen to challenge →
read the original abstract
Some argue scale is all what is needed to achieve AI, covering even causal models. We make it clear that large language models (LLMs) cannot be causal and give reason onto why sometimes we might feel otherwise. To this end, we define and exemplify a new subgroup of Structural Causal Model (SCM) that we call meta SCM which encode causal facts about other SCM within their variables. We conjecture that in the cases where LLM succeed in doing causal inference, underlying was a respective meta SCM that exposed correlations between causal facts in natural language on whose data the LLM was ultimately trained. If our hypothesis holds true, then this would imply that LLMs are like parrots in that they simply recite the causal knowledge embedded in the data. Our empirical analysis provides favoring evidence that current LLMs are even weak `causal parrots.'
This paper has not been read by Pith yet.
Forward citations
Cited by 9 Pith papers
-
YoCausal: How Far is Video Generation from World Model? A Causality Perspective
YoCausal benchmark shows video diffusion models detect the arrow of time but lack genuine causal understanding relative to humans.
-
Simulate, Reason, Decide: Scientific Reasoning with LLMs for Simulation-Driven Decision Making
MechSim is a mechanism-grounded framework that represents simulators via structured schemas and uses constrained LLM agents to generate evidence-based explanations linking outcomes to underlying mechanisms.
-
ORCA: An End-to-End Interactive Copilot for Optimized Root Cause Analysis
ORCA is an agent-orchestrated interactive copilot that automates and guides end-to-end causal analysis from workflow selection to report generation across real-world use cases.
-
CIVeX: Causal Intervention Verification for Language Agents
CIVeX maps agent tool calls to structural causal queries, checks identifiability, and issues auditable verdicts to prevent false executions while preserving utility on confounded benchmarks.
-
CauSim: Scaling Causal Reasoning with Increasingly Complex Causal Simulators
CauSim turns scarce causal reasoning labels into scalable supervised data by having LLMs incrementally construct complex executable structural causal models.
-
CogInstrument: Modeling Cognitive Processes for Bidirectional Human-LLM Alignment in Planning Tasks
CogInstrument represents human reasoning as revisable cognitive motifs in graphical form to support iterative alignment with LLMs during planning tasks, with a N=12 study indicating gains in targeted revision, agency,...
-
CounterBench: Evaluating and Improving Counterfactual Reasoning in Large Language Models
Introduces CounterBench benchmark and CoIn iterative reasoning method showing LLMs perform near random on formal counterfactual tasks but improve substantially with guided backtracking.
-
Novelty-based Tree-of-Thought Search for LLM Reasoning and Planning
Novelty estimation via LLM prompts enables pruning in Tree-of-Thought search, reducing overall token usage on language planning benchmarks.
-
Does the Model Say What the Data Says? A Simple Heuristic for Model Data Alignment
A data-derived baseline using feature effects on binary outcomes provides a model-agnostic way to check if machine learning explanations align with the underlying data structure.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.