pith. sign in

arxiv: 2502.04382 · v3 · pith:75UUPW26new · submitted 2025-02-05 · 💻 cs.CL · cs.AI· cs.CY

Sparse Autoencoders for Hypothesis Generation

classification 💻 cs.CL cs.AIcs.CY
keywords hypothesaestargetvariabledatadatasetsfeaturesheadlineshypotheses
0
0 comments X
read the original abstract

We describe HypotheSAEs, a general method to hypothesize interpretable relationships between text data (e.g., headlines) and a target variable (e.g., clicks). HypotheSAEs has three steps: (1) train a sparse autoencoder on text embeddings to produce interpretable features describing the data distribution, (2) select features that predict the target variable, and (3) generate a natural language interpretation of each feature (e.g., "mentions being surprised or shocked") using an LLM. Each interpretation serves as a hypothesis about what predicts the target variable. Compared to baselines, our method better identifies reference hypotheses on synthetic datasets (at least +0.06 in F1) and produces more predictive hypotheses on real datasets (~twice as many significant findings), despite requiring 1-2 orders of magnitude less compute than recent LLM-based methods. HypotheSAEs also produces novel discoveries on two well-studied tasks: explaining partisan differences in Congressional speeches and identifying drivers of engagement with online headlines.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. VFUSE: Virulent Feature Understanding with Sparse autoEncoders

    cs.LG 2026-06 unverdicted novelty 7.0

    VFUSE applies sparse autoencoders to diffusion-transformer activations in RoseTTAFold3 and RFDiffusion3 to find monosemantic features that detect hazardous protein designs with AUROC up to 0.84.

  2. Localizing RL-Induced Tool Use to a Single Crosscoder Feature

    cs.LG 2026-06 unverdicted novelty 6.0

    Dedicated Feature Crosscoders localize RL-induced tool use to a compact feature set in Qwen2.5-3B, yielding +31.1 pp tool correctness gains and +6.8 pp spillover to the base model.

  3. Making Interpretable Discoveries from Unstructured Data: A High-Dimensional Multiple Hypothesis Testing Approach

    econ.EM 2025-11 unverdicted novelty 6.0

    A new framework combines AI-derived concept embeddings with high-dimensional selective inference to enable statistically principled, interpretable discovery from unstructured data in empirical economics.

  4. Steered Generation via Gradient-Based Optimization on Sparse Query Features

    cs.LG 2026-05 unverdicted novelty 5.0

    Prototype-Based Sparse Steering decomposes query activations with SAEs and optimizes sparse features via gradients to steer LLM outputs toward specific behaviors.

  5. Position: Behavioural Assurance Cannot Verify the Safety Claims Governance Now Demands

    cs.LG 2026-05 unverdicted novelty 5.0

    Behavioral assurance is structurally unable to verify the latent safety properties demanded by AI governance frameworks enacted 2019-2026.