Lune

ACL Findings2026

SAE-FiRE: Enhancing Earnings Surprise Predictions Through Sparse Autoencoder Feature Selection

Huopu Zhang, Yanguang Liu, Miao Zhang, Zirui He, Mengnan Du

Abstract

Predicting earnings surprises from financial documents, such as earnings conference calls, regulatory filings, and financial news, has become increasingly important in financial economics. However, these financial documents present significant analytical challenges, typically containing over 5,000 words with substantial redundancy and industry-specific terminology that creates obstacles for language models. In this work, we propose the SAE-FiRE (Sparse Autoencoder for Financial Representation Enhancement) framework to address these limitations by extracting key information while eliminating redundancy. SAE-FiRE employs Sparse Autoencoders (SAEs) to decompose dense neural representations from large language models into interpretable sparse components, then applies statistical feature selection methods, including ANOVA F-tests and tree-based importance scoring, to identify the top-k most discriminative dimensions for classification. By systematically filtering out noise that might otherwise lead to overfitting, we enable more robust and generalizable predictions. Experimental results across three financial datasets demonstrate that SAE-FiRE significantly outperforms baseline approaches.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

1 top-tier paper has cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Do I Know This Entity? Knowledge Awareness and Hallucinations in Language Models cites Scaling and evaluating sparse autoencoders
  • FinCall-Surprise: A Large Scale Multi-modal Benchmark for Earning Surprise Prediction cites Forecasting Earnings Surprises from Conference Call Transcripts
  • Forecasting Earnings Surprises from Conference Call Transcripts cites Multimodal Multi-Task Financial Risk Forecasting
  • A Survey on Sparse Autoencoders: Interpreting the Internal Mechanisms of Large Language Models cites Do I Know This Entity? Knowledge Awareness and Hallucinations in Language Models
  • A Survey on Sparse Autoencoders: Interpreting the Internal Mechanisms of Large Language Models cites Scaling and evaluating sparse autoencoders
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers1

  1. FinCall-Surprise: A Large Scale Multi-modal Benchmark for Earning Surprise PredictionDong Shu, Yanguang Liu, Huopu Zhang, Mengnan DuACL 2026 · 1 citation

Builds on5

  1. Multimodal Multi-Task Financial Risk ForecastingRamit Sawhney, Puneet Mathur, Ayush Mangal, Piyush Khanna et al.ACM MM 2020 · 61 citations
  2. A Survey on Sparse Autoencoders: Interpreting the Internal Mechanisms of Large Language ModelsDong Shu, Xuansheng Wu, Haiyan Zhao, Daking Rai et al.EMNLP Findings 2025 · 13 citations
  3. Forecasting Earnings Surprises from Conference Call TranscriptsRoss Koval, Nicholas Andrews, Xifeng YanACL Findings 2023 · 10 citations
  4. Scaling and evaluating sparse autoencodersLeo Gao, Tom Dupré la Tour, Henk Tillman, Gabriel Goh et al.ICLR 2025 · 10 citations
  5. Do I Know This Entity? Knowledge Awareness and Hallucinations in Language ModelsJavier Ferrando, Oscar Balcells Obeso, Senthooran Rajamanoharan, Neel NandaICLR 2025 · 1 citation

Related papers

  1. Learning Retrieval Models with Sparse AutoencodersThibault Formal, Maxime Louis, Hervé Déjean, Stéphane ClinchantICLR 2026 · 9 citations
  2. Sparse Autoencoder Features for Classifications and TransferabilityJack Gallifant, Shan Chen, Kuleen Sasse, Hugo J. W. L. Aerts et al.EMNLP 2025
  3. A is for Absorption: Studying Feature Splitting and Absorption in Sparse AutoencodersDavid Chanin, James Wilken-Smith, Tomás Dulka, Hardik Bhatnagar et al.NeurIPS 2025 · 168 citations
  4. Rethinking Evaluation of Sparse Autoencoders through the Representation of Polysemous WordsGouki Minegishi, Hiroki Furuta, Yusuke Iwasawa, Yutaka MatsuoICLR 2025
  5. Selective Deep Autoencoder for Unsupervised Feature SelectionWael Hassanieh, Abdallah A. ChehadeAAAI 2024 · 16 citations
Dusk over the sea between two cliffs drawn in fine vertical lines