Interpretable Discovery from Unstructured Data: A High-Dimensional Approach

arXiv:2511.01680 · econ.EM, cs.LG · Submitted 2025-11-03 · Read on arXiv

econ.EM, cs.LG

Submitted: 2025-11-03

Updated: 2026-10-02

Code: https://github.com/jscarlson/hdmht-discovery

Terminology

• Embeddings — 2× in this paper · explained in 2 episodes
• Interpretability — 1× in this paper · explained in 15 episodes
• Post-Selection — 1× in this paper · explained in 2 episodes

Sources

• Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE)
• Interpretable Embeddings with Sparse Autoencoders: A Data Analysis Toolkit
• Automatically Interpreting Millions of Features in Large Language Models
• Causal Inference on Outcomes Learned from Text
• High-Dimensional Econometrics and Regularized GMM
• Large Language Models: An Applied Econometric Framework
• A Unifying Framework for Robust and Efficient Inference with Unstructured Data
• Program Evaluation with Remotely Sensed Outcomes
• Inference for Regression with Variables Generated by AI or Machine Learning
• Machine-Learning Tests for Effects on Multiple Outcomes
• Gaussian Multiplier Bootstrap Procedure for the $k$th Largest Coordinate of High-Dimensional Statistics
• High Dimensional Bootstrap and Asymptotic Expansion for the $k$-th Largest Coordinate
• Goal Driven Discovery of Distributional Differences via Language Descriptions
• Position: Use Sparse Autoencoders to Discover Unknowns
• Towards A Rigorous Science of Interpretable Machine Learning
• Sparse Autoencoders Do Not Find Canonical Units of Analysis
• Transcoders Beat Sparse Autoencoders for Interpretability
• A Multimodal Automated Interpretability Agent
• Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders
• An X-Ray Is Worth 15 Features: Sparse Autoencoders for Interpretable Radiology Report Generation

Related papers

• SLIM: Stochastic Learning and Inference in Overidentified Models
• High-dimensional censored MIDAS logistic regression for corporate survival forecasting
• Cross-Fitting-Free Debiased Machine Learning with Multiway Dependence
• Policy design in experiments with unknown interference
• Can large language models assist choice modelling? Insights into prompting strategies and current models' capabilities
• Causal Inference in Possibly Nonlinear Factor Models