SOURCE-LINKED INTELLIGENCE

Exploring Sparse Autoencoders in Text-Based Causal Confounding Adjustment

arXiv · AI, language, vision and robotics · article · Sep 1, 2026 · UTC

In many settings, studying causal questions based on text data requires adjusting for confounding information within texts. Yet there is a tradeoff in constructing text representations for adjustment: they must be sufficiently large and/or dense to preserve the confounding variables necessary for unbiased effect estimation, but sufficiently small and/or sparse to satisfy finite-sample overlap and yield low-variance estimates. To address this tradeoff, we turn to sparse autoencoders (SAEs), and propose a novel causal adjustment pipeline that iteratively selects a minimal set of SAE features via

Read original source ↗ Open in workspace

recordType

paper

region

Global