---
title: "Actionable Interpretability for Churn Classification: A Text"
description: "This industry case study, carried out by the University of Amsterdam and KPN, applies the Text Bottleneck Model (TBM) of… 30 references."
authors: ["Adrian Sauter", "Vera Neplenbroek", "Georgios Vlassopoulos", "Gianluigi Bardelloni"]
year: 2026
venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)"
source: "https://aclanthology.org/2026.acl-industry.70/"
doi: "10.18653/v1/2026.acl-industry.70"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/sauter-2026-actionable-interpretability-churn-classification-text/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/sauter-2026-actionable-interpretability-churn-classification-text/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/21/) / Actionable Interpretability for Churn Classification: A Text Bottleneck Model Case Study at a Major Telecom Provider

# Actionable Interpretability for Churn Classification: A Text Bottleneck Model Case Study at a Major Telecom Provider

**Authors:** Adrian Sauter, [Vera Neplenbroek](https://md.research.elamur.ai/free/en/authors/vera-neplenbroek/), Georgios Vlassopoulos and Gianluigi Bardelloni · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2026](https://md.research.elamur.ai/free/en/venues/acl/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.acl-industry.70/) · [PDF](https://aclanthology.org/2026.acl-industry.70.pdf) · [DOI 10.18653/v1/2026.acl-industry.70](https://doi.org/10.18653/v1/2026.acl-industry.70) · [OpenAlex](https://openalex.org/W7166809154)

**Citations:** 0 citations (OpenAlex, 2026-10-14)

## Summary
Telecom operators need to know why a customer wants to cancel, not only whether they do. Black-box LLM classifiers return only a label, which gives retention teams nothing to act on. This industry case study, carried out by the University of Amsterdam and KPN, applies the Text Bottleneck Model (TBM) of Ludan et al. (2024) to churn classification after customer service calls.

The pipeline has three steps:
- An LLM proposes candidate concepts, such as Price Sentiment or Provider Switch, and domain experts refine them.
- An LLM rates every transcript on each concept and extracts evidence snippets.
- A white-box classifier, such as logistic regression, predicts Churn/No Churn from the resulting concept vector.

Experiments use a small expert-annotated set of call transcripts and GPT-4 models. The expert-refined TBM performs on par with black-box baselines. Its concept activations also support clustering-based call profiling and an interactive dashboard for non-technical managers.

The work suggests that interpretability need not cost predictive power in an industrial NLP setting. Steps toward deployment are reportedly underway.

## Key points
- **Data:** Each transcript was labeled by five annotators under a 14-class schema. "Cancel subscription", including partial cancellations or downgrades, maps to Churn.
  - Training set: 200 transcripts, balanced 100/100.
  - Test set: 200 transcripts with the natural imbalance (28 Churn, 172 No Churn).
  - Transcripts average about 270 words and 20 turns. All examples shown in the paper are synthetic.
- **LLMs:** Concept generation and measurement use gpt-4-turbo-128k, with gpt-4o-mini tested as a cheaper option.
- **Baselines:**
  - A BERTje classifier, domain-adapted on 1.5M dialogues and limited to the first 512 tokens.
  - Direct GPT prompting.
- **Automated concept generation:** The acceptance threshold was γ = 0.1. Three runs with 8 concepts each reached test accuracies between 0.64 and 0.72.
- **Call profiling:** k-means clustering on concept vectors (four clusters) separated call profiles, for example a cluster with a 98.1% churn rate versus one with 15.6%.
- **Limitations:**
  - No human evaluation of concept quality or measurements.
  - High cost, since each concept needs a separate prompt; batching concepts into one prompt degraded measurement quality.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [University of Amsterdam](https://md.research.elamur.ai/free/en/universities/university-of-amsterdam/) — University, Amsterdam, Netherlands · 144 works · 173,906 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [💼 Business, Management and Accounting](https://md.research.elamur.ai/free/en/shelves/business-management-and-accounting/) · 32 works · [💼 Marketing](https://md.research.elamur.ai/free/en/shelves/business-management-and-accounting/marketing/) · 4 works

🧭 Topics (1): [🔬 Interpretability](https://md.research.elamur.ai/free/en/topics/interpretability/) · 130 works

## Cite

Cite as: Adrian Sauter, Vera Neplenbroek, Georgios Vlassopoulos and Gianluigi Bardelloni. 2026. Actionable Interpretability for Churn Classification: A Text Bottleneck Model Case Study at a Major Telecom Provider. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{sauter2026actionable,
  title = {{Actionable Interpretability for Churn Classification: A Text Bottleneck Model Case Study at a Major Telecom Provider}},
  author = {Adrian Sauter, Vera Neplenbroek, Georgios Vlassopoulos and Gianluigi Bardelloni},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-industry.70},
  url = {https://doi.org/10.18653/v1/2026.acl-industry.70}
}
```

- [← previous Action Boundary Blindness: When LLM Agents Cannot Tell Where One Action Ends and Another Begins](https://md.research.elamur.ai/free/en/papers/wang-2026-action-boundary-blindness-when-llm/)
- [next → Actors, Frames and Arguments: A Multi-Decade Computational Analysis of Climate Discourse in Financial News using Large Language Models](https://md.research.elamur.ai/free/en/papers/su-2026-actors-frames-arguments-multi-decade/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/21/)

## References

30 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Eldar David Abraham, Karel D’Oosterlinck, Amir Feder, Yair Ori Gat, Atticus Geiger, Christopher Potts, Roi Reichart, and Zhengxuan Wu. 2022. CE-Bab: Estimating the causal effects of real-world concepts on NLP model behavior. In Advances in Neural Information Processing Systems. [find](https://scholar.google.com/scholar?q=Eldar+David+Abraham%2C+Karel+D%E2%80%99Oosterlinck%2C+Amir+Feder%2C+Yair+Ori+Gat%2C+Atticus+Geiger%2C+Christopher+Potts%2C+Roi+Reichart%2C+and+Zhengxuan+Wu%2E+2022%2E+CE%2DBab%3A+Estimating+the+causal+effects+of+real%2Dworld+concept)
2. Julius Adebayo, Justin Gilmer, Michael Muelly, Ian Goodfellow, Moritz Hardt, and Been Kim. 2018. Sanity checks for saliency maps. In Advances in Neural Information Processing Systems, volume 31. Curran Associates, Inc. [find](https://scholar.google.com/scholar?q=Julius+Adebayo%2C+Justin+Gilmer%2C+Michael+Muelly%2C+Ian+Goodfellow%2C+Moritz+Hardt%2C+and+Been+Kim%2E+2018%2E+Sanity+checks+for+saliency+maps%2E+In+Advances+in+Neural+Information+Processing+Systems%2C+volume+31%2E+Curra)
3. Jasmijn Bastings, Wilker Aziz, and Ivan Titov. 2019. [Interpretable neural predictions with differentiable binary variables](https://md.research.elamur.ai/free/en/papers/bastings-2019-interpretable-neural-predictions-differentiable-binary/). In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 2963–2977, Florence, Italy. Association for Computational Linguistics.
4. Anna Bavaresco, Raffaella Bernardi, Leonardo Bertolazzi, Desmond Elliott, Raquel Fernández, Albert Gatt, Esam Ghaleb, Mario Giulianelli, Michael Hanna, Alexander Koller, Andre Martins, Philipp Mondorf, Vera Neplenbroek, Sandro Pezzelle, Barbara Plank, David Schlangen, Alessandro Suglia, Aditya K Sur… — [LLMs instead of Human Judges? A Large Scale Empirical Study across 20 NLP Evaluation Tasks](https://md.research.elamur.ai/free/en/papers/bavaresco-2024-llms-instead-human-judges-large/)
5. Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel Ziegler, Jeffrey Wu, Clemens Winter, and… [find](https://scholar.google.com/scholar?q=Tom+Brown%2C+Benjamin+Mann%2C+Nick+Ryder%2C+Melanie+Subbiah%2C+Jared+D+Kaplan%2C+Prafulla+Dhariwal%2C+Arvind+Neelakantan%2C+Pranav+Shyam%2C+Girish+Sastry%2C+Amanda+Askell%2C+Sandhini+Agarwal%2C+Ariel+Herbert%2DVoss%2C+Gretchen)
6. Joseph L Fleiss. 1971. Measuring nominal scale agreement among many raters. Psychological bulletin, 76(5):378. [find](https://scholar.google.com/scholar?q=Joseph+L+Fleiss%2E+1971%2E+Measuring+nominal+scale+agreement+among+many+raters%2E+Psychological+bulletin%2C+76%285%29%3A378%2E)
7. Leilani H. Gilpin, David Bau, Ben Z. Yuan, Ayesha Bajwa, Michael Specter, and Lalana Kagal. 2018. Explaining explanations: An overview of interpretability of machine learning. In 2018 IEEE 5th International Conference on Data Science and Advanced Analytics (DSAA), pages 80–89. [find](https://scholar.google.com/scholar?q=Leilani+H%2E+Gilpin%2C+David+Bau%2C+Ben+Z%2E+Yuan%2C+Ayesha+Bajwa%2C+Michael+Specter%2C+and+Lalana+Kagal%2E+2018%2E+Explaining+explanations%3A+An+overview+of+interpretability+of+machine+learning%2E+In+2018+IEEE+5th+Interna)
8. Vincent Huang, Dami Choi, Daniel D. Johnson, Sarah Schwettmann, and Jacob Steinhardt. 2025. Predictive concept decoders: Training scalable end-to-end interpretability assistants. Preprint, arXiv:2512.15712. [find](https://scholar.google.com/scholar?q=Vincent+Huang%2C+Dami+Choi%2C+Daniel+D%2E+Johnson%2C+Sarah+Schwettmann%2C+and+Jacob+Steinhardt%2E+2025%2E+Predictive+concept+decoders%3A+Training+scalable+end%2Dto%2Dend+interpretability+assistants%2E+Preprint%2C+arXiv%3A2512%2E)
9. Alon Jacovi and Yoav Goldberg. 2020. [Towards faithfully interpretable NLP systems: How should we define and evaluate faithfulness?](https://md.research.elamur.ai/free/en/papers/jacovi-2020-faithfully-interpretable-nlp-systems-how/) In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 4198–4205, Online. Association for Computational Linguistics.
10. Sarthak Jain and Byron C. Wallace. 2019. [Attention is not Explanation](https://md.research.elamur.ai/free/en/papers/jain-2019-attention-is-not-explanation/). In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), pages 3543–3556, Minneapolis, Minnesota. Associa…
11. Albert Q. Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, Lélio Renard Lavaud, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix,… [find](https://scholar.google.com/scholar?q=Albert+Q%2E+Jiang%2C+Alexandre+Sablayrolles%2C+Arthur+Mensch%2C+Chris+Bamford%2C+Devendra+Singh+Chaplot%2C+Diego+de+las+Casas%2C+Florian+Bressand%2C+Gianna+Lengyel%2C+Guillaume+Lample%2C+Lucile+Saulnier%2C+L%C3%A9lio+Renard+Lav)
12. Pang Wei Koh, Thao Nguyen, Yew Siang Tang, Stephen Mussmann, Emma Pierson, Been Kim, and Percy Liang. 2020. Concept bottleneck models. In Proceedings of the 37th International Conference on Machine Learning, volume 119 of Proceedings of Machine Learning Research, pages 5338–5348. PMLR. [find](https://scholar.google.com/scholar?q=Pang+Wei+Koh%2C+Thao+Nguyen%2C+Yew+Siang+Tang%2C+Stephen+Mussmann%2C+Emma+Pierson%2C+Been+Kim%2C+and+Percy+Liang%2E+2020%2E+Concept+bottleneck+models%2E+In+Proceedings+of+the+37th+International+Conference+on+Machine+Le)
13. Ryan Koo, Minhwa Lee, Vipul Raheja, Jong Inn Park, Zae Myung Kim, and Dongyeop Kang. 2024. Benchmarking cognitive biases in large language models as evaluators. In Findings of the Association for Computational Linguistics: ACL 2024, pages 517–545, Bangkok, Thailand. Association for Computational Lin… [find](https://scholar.google.com/scholar?q=Ryan+Koo%2C+Minhwa+Lee%2C+Vipul+Raheja%2C+Jong+Inn+Park%2C+Zae+Myung+Kim%2C+and+Dongyeop+Kang%2E+2024%2E+Benchmarking+cognitive+biases+in+large+language+models+as+evaluators%2E+In+Findings+of+the+Association+for+Comp)
14. Chaofan Chen, Kangcheng Lin, Cynthia Rudin, Yaron Shaposhnik, Sijia Wang, and Tong Wang. 2018. An interpretable model with globally consistent explanations for credit risk. Preprint, arXiv:1811.12615. [find](https://scholar.google.com/scholar?q=Chaofan+Chen%2C+Kangcheng+Lin%2C+Cynthia+Rudin%2C+Yaron+Shaposhnik%2C+Sijia+Wang%2C+and+Tong+Wang%2E+2018%2E+An+interpretable+model+with+globally+consistent+explanations+for+credit+risk%2E+Preprint%2C+arXiv%3A1811%2E12615%2E)
15. Tao Lei, Regina Barzilay, and Tommi Jaakkola. 2016. [Rationalizing neural predictions](https://md.research.elamur.ai/free/en/papers/lei-2016-rationalizing-neural-predictions/). In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 107–117, Austin, Texas. Association for Computational Linguistics.
16. Wietse de Vries, Andreas van Cranenburgh, Arianna Bisazza, Tommaso Caselli, Gertjan van Noord, and Malvina Nissim. 2019. [BERTje: A Dutch BERT Model](https://md.research.elamur.ai/free/en/papers/de-vries-2019-bertje-dutch-bert-model/). arXiv:1912.09582.
17. Percy Liang, Rishi Bommasani, Tony Lee, Dimitris Tsipras, Dilara Soylu, Michihiro Yasunaga, Yian Zhang, Deepak Narayanan, Yuhuai Wu, Ananya Kumar, Benjamin Newman, Binhang Yuan, Bobby Yan, Ce Zhang, Christian Cosgrove, Christopher D Manning, Christopher Re, Diana Acosta-Navas, Drew A. Hudson, and 31… [find](https://scholar.google.com/scholar?q=Percy+Liang%2C+Rishi+Bommasani%2C+Tony+Lee%2C+Dimitris+Tsipras%2C+Dilara+Soylu%2C+Michihiro+Yasunaga%2C+Yian+Zhang%2C+Deepak+Narayanan%2C+Yuhuai+Wu%2C+Ananya+Kumar%2C+Benjamin+Newman%2C+Binhang+Yuan%2C+Bobby+Yan%2C+Ce+Zhang%2C+C)
18. Josh Magnus Ludan, Qing Lyu, Yue Yang, Liam Dugan, Mark Yatskar, and Chris Callison-Burch. 2024. Interpretable-by-design text understanding with iteratively generated concept bottleneck. Preprint, arXiv:2310.19660. [find](https://scholar.google.com/scholar?q=Josh+Magnus+Ludan%2C+Qing+Lyu%2C+Yue+Yang%2C+Liam+Dugan%2C+Mark+Yatskar%2C+and+Chris+Callison%2DBurch%2E+2024%2E+Interpretable%2Dby%2Ddesign+text+understanding+with+iteratively+generated+concept+bottleneck%2E+Preprint%2C+arX)
19. Scott M. Lundberg and Su-In Lee. 2017. [A unified approach to interpreting model predictions](https://md.research.elamur.ai/free/en/papers/lundberg-2017-unified-approach-interpreting-model-predictions/). In Proceedings of the 31st International Conference on Neural Information Processing Systems, NIPS’17, page 4768–4777, Red Hook, NY, USA. Curran Associates Inc.
20. OpenAI. 2024. Gpt-4o model card. [find](https://scholar.google.com/scholar?q=OpenAI%2E+2024%2E+Gpt%2D4o+model+card%2E)
21. Dheeraj Rajagopal, Vidhisha Balachandran, Eduard H Hovy, and Yulia Tsvetkov. 2021. SELFEXPLAIN: A self-explaining architecture for neural text classifiers. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 836– 850, Online and Punta Cana, Dominican Repu… [find](https://scholar.google.com/scholar?q=Dheeraj+Rajagopal%2C+Vidhisha+Balachandran%2C+Eduard+H+Hovy%2C+and+Yulia+Tsvetkov%2E+2021%2E+SELFEXPLAIN%3A+A+self%2Dexplaining+architecture+for+neural+text+classifiers%2E+In+Proceedings+of+the+2021+Conference+on+Emp)
22. Marco Tulio Ribeiro, Sameer Singh, and Carlos Guestrin. 2016. ["why should i trust you?": Explaining the predictions of any classifier](https://md.research.elamur.ai/free/en/papers/ribeiro-2016-why-should-i-trust-explaining/). In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD ’16, page 1135–1144, New York, NY, USA. Association for Co…
23. Cynthia Rudin. 2019. [Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead](https://md.research.elamur.ai/free/en/papers/rudin-2018-stop-explaining-black-box/). Nature machine intelligence, 1(5):206–215.
24. Emily Sheng and David Uthus. 2020. Investigating societal biases in a poetry composition system. In Proceedings of the Second Workshop on Gender Bias in Natural Language Processing, pages 93–106, Barcelona, Spain (Online). Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Emily+Sheng+and+David+Uthus%2E+2020%2E+Investigating+societal+biases+in+a+poetry+composition+system%2E+In+Proceedings+of+the+Second+Workshop+on+Gender+Bias+in+Natural+Language+Processing%2C+pages+93%E2%80%93106%2C+Barc)
25. Chung-En Sun, Tuomas Oikarinen, Berk Ustun, and Tsui-Wei Weng. 2025. Concept bottleneck large language models. In The Thirteenth International Conference on Learning Representations. [find](https://scholar.google.com/scholar?q=Chung%2DEn+Sun%2C+Tuomas+Oikarinen%2C+Berk+Ustun%2C+and+Tsui%2DWei+Weng%2E+2025%2E+Concept+bottleneck+large+language+models%2E+In+The+Thirteenth+International+Conference+on+Learning+Representations%2E)
26. Zhen Tan, Lu Cheng, Song Wang, Bo Yuan, Jundong Li, and Huan Liu. 2024. Interpreting pretrained language models via concept bottlenecks. In Advances in Knowledge Discovery and Data Mining, pages 56–74, Singapore. Springer Nature Singapore. [find](https://scholar.google.com/scholar?q=Zhen+Tan%2C+Lu+Cheng%2C+Song+Wang%2C+Bo+Yuan%2C+Jundong+Li%2C+and+Huan+Liu%2E+2024%2E+Interpreting+pretrained+language+models+via+concept+bottlenecks%2E+In+Advances+in+Knowledge+Discovery+and+Data+Mining%2C+pages+56%E2%80%9374)
27. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, brian ichter, Fei Xia, Ed H. Chi, Quoc V Le, and Denny Zhou. 2022. [Chain of thought prompting elicits reasoning in large language models](https://md.research.elamur.ai/free/en/papers/wei-2022-chain-thought-prompting-elicits-reasoning/). In Advances in Neural Information Processing Systems.
28. Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chaumond, Clement Delangue, Anthony Moi, Pier-ric Cistac, Tim Rault, Remi Louf, Morgan Funtowicz, Joe Davison, Sam Shleifer, Patrick von Platen, Clara Ma, Yacine Jernite, Julien Plu, Canwen Xu, Teven Le Scao, Sylvain Gugger, and 3 others. 2020. [Transf…](https://md.research.elamur.ai/free/en/papers/wolf-2020-transformers-state-art-natural-language/)
29. Haiyan Zhao, Hanjie Chen, Fan Yang, Ninghao Liu, Huiqi Deng, Hengyi Cai, Shuaiqiang Wang, Dawei Yin, and Mengnan Du. 2024. [Explainability for large language models: A survey](https://md.research.elamur.ai/free/en/papers/zhao-2023-explainability-large-language-models-survey/). ACM Trans. Intell. Syst. Technol., 15(2).
30. Chunting Zhou, Pengfei Liu, Puxin Xu, Srini Iyer, Jiao Sun, Yuning Mao, Xuezhe Ma, Avia Efrat, Ping Yu, LILI YU, Susan Zhang, Gargi Ghosh, Mike Lewis, Luke Zettlemoyer, and Omer Levy. 2023. LIMA: Less is more for alignment. In Thirty-seventh Conference on Neural Information Processing Systems. [find](https://scholar.google.com/scholar?q=Chunting+Zhou%2C+Pengfei+Liu%2C+Puxin+Xu%2C+Srini+Iyer%2C+Jiao+Sun%2C+Yuning+Mao%2C+Xuezhe+Ma%2C+Avia+Efrat%2C+Ping+Yu%2C+LILI+YU%2C+Susan+Zhang%2C+Gargi+Ghosh%2C+Mike+Lewis%2C+Luke+Zettlemoyer%2C+and+Omer+Levy%2E+2023%2E+LIMA%3A+Less)

*Updated 1 November 2026*