causality

3 items

RESEARCHarXiv CS.AI·22d ago

Fair outputs, Biased Internals: Causal Potency and Asymmetry of Latent Bias in LLMs for High-Stakes Decisions

This research paper explores the disconnect between fair outputs of language models and their latent internal biases in high-stakes decisions like mortgage underwriting. It demonstrates that while LLMs may show no output bias, they retain and amplify demographic representations which can cause decision reversals, and this bias is asymmetric.

LLM bias machine learning causality AI ethics

ARTICLEDEV.to AI·20d ago

Judea Pearl's Ladder of Causation and the Limits of LLM Reasoning

This article explores the fundamental limitations of Large Language Models (LLMs) in causal reasoning, referencing Judea Pearl's Ladder of Causation. It highlights that LLMs often operate at the lowest rung of association, failing to identify true causes and instead patching correlations, which explains common errors in AI tools.

AI limitations Judea Pearl causality AI Reasoning

RESEARCHDEV.to AI·4/11/2026

Learning Neural Causal Models from Unknown Interventions

This work explores methods for neural models to learn cause-and-effect relationships, even in scenarios where data-generating interventions are unknown. The research aims to enhance artificial intelligence's ability to infer causality from complex data.

neural networks machine learning causality causal inference