Pith. sign in

REVIEW 4 cited by

Interpretable-by-Design Text Understanding with Iteratively Generated Concept Bottleneck

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2310.19660 v2 pith:VCMTXYZE submitted 2023-10-30 cs.CL

Interpretable-by-Design Text Understanding with Iteratively Generated Concept Bottleneck

classification cs.CL
keywords textconceptconceptshumanblack-boxbottleneckclassificationfinetuned
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Black-box deep neural networks excel in text classification, yet their application in high-stakes domains is hindered by their lack of interpretability. To address this, we propose Text Bottleneck Models (TBM), an intrinsically interpretable text classification framework that offers both global and local explanations. Rather than directly predicting the output label, TBM predicts categorical values for a sparse set of salient concepts and uses a linear layer over those concept values to produce the final prediction. These concepts can be automatically discovered and measured by a Large Language Model (LLM) without the need for human curation. Experiments on 12 diverse text understanding datasets demonstrate that TBM can rival the performance of black-box baselines such as few-shot GPT-4 and finetuned DeBERTa while falling short against finetuned GPT-3.5. Comprehensive human evaluation validates that TBM can generate high-quality concepts relevant to the task, and the concept measurement aligns well with human judgments, suggesting that the predictions made by TBMs are interpretable. Overall, our findings suggest that TBM is a promising new framework that enhances interpretability with minimal performance tradeoffs.

discussion (0)

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

Forward citations

Cited by 4 Pith papers

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

  1. When, How Long and How Much? Interpretable Neural Networks for Time Series Regression by Learning to Mask and Aggregate

    cs.LG 2025-12 conditional novelty 7.0

    MAGNETS learns unsupervised, mask-based concepts to make time-series regression predictions additively interpretable, recovering ground-truth temporal rules on synthetic tasks and beating interpretable baselines on mo...

  2. Mitigating Label Bias with Interpretable Rubric Embeddings

    cs.LG 2026-05 unverdicted novelty 6.0

    Rubric embeddings from expert criteria mitigate label bias in models trained on historical evaluations, reducing group disparities while improving cohort quality on a master's program dataset.

  3. Interpretable Discriminative Text Representations via Agreement and Label Disentanglement

    cs.CL 2026-05 unverdicted novelty 6.0

    LFD discovers predictive text features via LLM contrastive proposals, cross-LLM Cohen's kappa screening, and residual held-out gain selection, matching baseline accuracy while achieving higher human agreement and lowe...

  4. Human-AI Co-design for Clinical Prediction Models

    cs.AI 2026-01 conditional novelty 5.0

    HACHI, an iterative human-AI loop that turns clinical notes into small interpretable yes/no-concept risk scores, outperforms standard instruments in two retrospective UCSF case studies.