Metraipolis › archive › dc_8d5fe97db23c
The Hidden States Cookbook: A Large-Scale Ablation Study for Noise-Robust Conversational Intent Classification in Industry — arXiv:2610.05394
- Source
- arXiv
- Original
- https://arxiv.org/abs/2610.05394
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-06T05:28:21.076Z
- SHA-256 of the archived text
a2110d340b1a88b2b029efeb8c1d175a21555b25a4a28b370f8361fa540a9d34- Stable URL of this page
https://metraipolis.com/documento/dc_8d5fe97db23c
How to cite
The Hidden States Cookbook: A Large-Scale Ablation Study for Noise-Robust Conversational Intent Classification in Industry — arXiv:2610.05394 — arXiv, https://arxiv.org/abs/2610.05394 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_8d5fe97db23c (Metraipolis archive dc_8d5fe97db23c, sha256 a2110d340b1a88b2b029efeb8c1d175a21555b25a4a28b370f8361fa540a9d34)
BibTeX
@misc{dc8d5fe97db23c,
title = {The Hidden States Cookbook: A Large-Scale Ablation Study for Noise-Robust Conversational Intent Classification in Industry — arXiv:2610.05394},
howpublished = {arXiv},
url = {https://arxiv.org/abs/2610.05394},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_8d5fe97db23c; sha256 a2110d340b1a88b2b029efeb8c1d175a21555b25a4a28b370f8361fa540a9d34},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.05394. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Conversational database interfaces face a critical challenge: users naturally embed queries in conversational noise (greetings, politeness, off-topic remarks), which degrades intent classification accuracy and wastes computational resources. Despite advances in orchestration and retrieval strategies, a fundamental question remains unanswered: which pooling strategy maximizes intent classification accuracy under realistic conversational noise in production language models? This work addresses this gap through 360 controlled experiments spanning four pooling configurations (mean, max, last-token, attention, and FFT-augmented variants) using Llama-3.2-1B-Instruct on BANKING77 and CLINC150 datasets under clean/noisy conditions with ten random seeds. Key findings reveal that attention pooling consistently outperforms alternative strategies under noisy conditions (~+2.6-2.8 F1 over the default), while mean pooling degrades performance by up to ~5 F1 points. Frequency-domain filtering does not produce consistent accuracy improvements and functions primarily as a structural variation rather than an accuracy-enhancing component. These results provide concrete, evidence-based guidance for building noise-robust conversational classifiers: attention pooling is recommended for noisy interfaces, mean pooling should be avoided, and last-token pooling is appropriate for clean-query scenarios. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.CL, nuovi lavori del giorno). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2610.05394 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-06T05:24:20.333Z dalla pagina https://arxiv.org/list/cs.CL/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).