Pith. sign in

REVIEW 4 major objections 5 minor 50 references

Transforming NLU with Babylon: A Case Study in Development of Real-time, Edge-Efficient, Multi-Intent Translation System for Automated Drive-Thru Ordering

T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read Babylon translates noisy ASR phonemes into structured transcode intents with 90.07% exact-match accuracy, 83 ms per order turn, and 118 MB memory, beating much larger NMT baselines.

desk verdict Babylon is a plausible edge-NLU architecture, but the paper's headline accuracy-latency-memory superiority over Flan-T5/BART is unmeasured because Table 2 reports no accuracy for those baselines. read the letter →

arxiv 2411.15372 v1 pith:4EBDBIPX submitted 2024-11-22 cs.CL cs.AIcs.LG

classification cs.CLcs.AIcs.LG
keywords naturallanguageunderstandingintentdetectionslotfillingtranscodespeech-to-intenttranslationedgeinferencelow-latencyNLUdrive-thruordering
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that natural language understanding for noisy, real-time drive-thru ordering can be recast as a translation task: convert ASR phoneme sequences into sequences of 'transcodes,' a compact regular language whose tokens encode both intents and slot values. It introduces Babylon, a 7M-parameter transformer with an LSTM token-pooling front end that compresses input, preserves word order, and suppresses ASR noise. The authors report 90.07% exact-match accuracy on real drive-thru orders at 83 ms per turn and 118 MB memory on a single CPU core, while Flan-T5-small and BART-base are slower, larger, and too inaccurate on phoneme input to be usable. If the transfer from synthetic training data to real orders holds, the approach offers a practical template for edge-deployed conversational understanding in other noisy kiosk domains.

What carries the argument

The central object is the transcode, a token from a purpose-built regular language whose sequences encode intents such as add, delete, and update together with slot values like item, quantity, size, and extras. The load-bearing mechanism is Babylon's input front end: a bidirectional LSTM that reads phoneme sequences, followed by token pooling that keeps every k-th LSTM output (k=4), starting from the end so the EOS token is retained. This shrinks sequence length fourfold before the standard transformer encoder-decoder, cutting computation, preserving temporal order that transformers can miss, and averaging over noisy phonemes. The paper's argument is that this simple front end achieves the reported accuracy-latency-memory trade-off without chunked or windowed attention.

What would settle it

Run Babylon and a vanilla transformer with comparable training on the same held-out set of real drive-thru orders with verified transcode labels and injected ASR phoneme errors; if Babylon's exact-match accuracy does not stay near 90% while its latency and memory remain near 83 ms and 118 MB on a single CPU core, the central trade-off claim fails.

Watch

Extended reading notes

Core claim

The central claim is that representing spoken customer orders as sequences of regular-language units (transcodes) makes multi-intent NLU a sequence-to-sequence translation problem, and that Babylon's bidirectional LSTM plus token-pooling front end lets a small transformer solve it accurately and cheaply at the edge. The paper argues that feeding raw phonemes through an LSTM before pooling every fourth output produces a context-rich, downsampled sequence that the transformer can translate with 90.07% exact-match accuracy, 83 ms average latency per order turn, and a 118 MB footprint on a single CPU core. Against this, Flan-T5-small and BART-base hallucinate or misparse phoneme inputs and are too slow and memory-heavy for edge deployment; their accuracy is not reported because ASR errors cascade on word-level inputs. The LSTM is also claimed to act as a regularizer, making the model robust to erroneous phonemes from upstream ASR, with examples where a vanilla transformer drops or misbinds items that Babylon recovers.

Load-bearing premise

The load-bearing premise is that the 250 million synthetic training orders resemble real drive-thru speech closely enough in accents, dialects, repetition patterns, and ASR noise, and that the 5,587-order test set's transcode labels are correct; the paper does not document the generator or the labeling process, and both datasets are withheld for proprietary reasons.

Editorial extensions

If this is right

  • A 7M-parameter NLU component can run on a single CPU core in single-threaded mode, so drive-thru ordering systems need not depend on cloud connectivity or GPU hardware.
  • Multiple intents within one customer turn are resolved in a single inference call, allowing dialog management to update order state in one step rather than re-prompting.
  • Operating on phonemes rather than words makes the same model robust to imperfect ASR and, because the representation is IPA-based, potentially transferable to other spoken languages.
  • BLEU and ROUGE do not track customer-intent errors in transcode output; exact-match transcode accuracy is the metric that matters for this application.
  • The paper's architecture can be reused in other noise-prone edge scenarios such as ticketing kiosks, which it identifies as a natural extension.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper reports only k=4 pooling; varying k on the same data would map a pooling-rate versus latency-accuracy frontier, a sweep the authors leave for future work.
  • Because the input is phoneme-based and the output is a constrained transcode language, the same synthetic-data recipe could generate labeled training sets for other menu-driven domains without human annotation, which the paper does not claim explicitly.
  • If the synthetic-to-real transfer holds as reported, a testable extension is that adding controlled phoneme substitutions at training time should improve robustness further; the paper does not run such an ablation.
  • The hallucination and latency results for Flan-T5 and BART suggest that general-purpose instruction-tuned models may need explicit phoneme-to-transcode fine-tuning or constrained decoding before they can compete, which the paper only hints at.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes Babylon, a 7M-parameter transformer-based NLU model for automated drive-thru ordering. NLU is framed as a translation task from phoneme sequences produced by an upstream ASR into sequences of 'transcodes', a regular language that encodes intents and slot values. The main architectural contribution is a bidirectional LSTM followed by token pooling placed before a standard transformer encoder-decoder, intended to reduce sequence length, improve robustness to ASR errors, and lower latency and memory footprint on edge CPUs. The experiments compare Babylon against four in-house models (Atlantis, Delphi, Camelot) and two open-source NMT baselines (Flan-T5-small, BART-base) on a proprietary test set of 5,587 real drive-thru orders, reporting exact-match accuracy, per-order and per-phoneme latency, and memory footprint. The paper claims that Babylon achieves 'significantly better accuracy-latency-memory footprint trade-offs' over Flan-T5 and BART.

Significance. If fully supported, this would be a useful engineering contribution: a small edge-deployable model that handles multi-intent spoken orders from noisy phoneme input, with 90.07% exact-match accuracy, 83 ms average per-order latency on a single CPU core, and 118 MB memory footprint. The LSTM-plus-pooling preprocessing is simple and plausible, and the qualitative examples in Appendix A.3 illustrate failure modes that the architecture may genuinely address. However, the evidence as presented is not sufficient to establish the headline trade-off claim: the accuracy coordinate for the two named external baselines is missing, no uncertainty or significance testing is reported, the token-pooling stride appears to be selected on the same test set, and both the synthetic training dataset and the real test set are withheld and not described in enough detail for external validation. The paper is honest about the proprietary constraints and includes a useful qualitative error analysis, but the central comparison remains unquantified.

major comments (4)
  1. [Abstract; §6; Table 2] Table 2 reports accuracy as 'N/A*' for Flan-T5-small and BART-base, with the footnote citing cascading ASR errors. Yet the abstract and §6 assert that Babylon achieves 'significantly better accuracy-latency-memory footprint trade-offs over typically employed NMT models like Flan-T5 and BART'. Without an accuracy measurement for these baselines, the trade-off comparison has only two of its three coordinates populated, so the claim is not supported by the reported data. This is the paper's central claim, not a minor omission. Please report an accuracy metric (exact-match or a clearly defined relaxation) for the baselines, or explicitly restrict the head-to-head claim to latency and memory only.
  2. [§6; Table 2] The in-house accuracy differences are small: Babylon 90.07%, Atlantis 89.74%, Camelot 88.87%, and Delphi 88.84%, all point estimates from a single run on 5,587 test orders. No confidence intervals, standard deviations across seeds, or significance tests (e.g., McNemar's test) are provided. The sentence in §6 that the 1.2% lift over Camelot 'significantly affects the application's business metrics' is not supported by any reported statistics. A formal comparison with uncertainty estimates is needed before declaring Babylon the best-performing model on accuracy.
  3. [§4.2; §7] The token pooling stride k was selected by experimentation, with k=4 chosen 'for optimal performance'. No accuracy results for other values of k are reported, and the paper does not state whether the selection was made on a held-out validation set or on the test set used for the headline numbers. If k was tuned on the test set, the reported 90.07% accuracy is an optimistically biased estimate. Please report the k-sweep results or clarify the selection procedure, and state which split was used for model selection.
  4. [§5.2; footnote in §5.2] The training set is 250 million synthetic orders generated by the method of Arel et al. (2019), a US patent, but the generation procedure is not described beyond the citation. The test set (5,587 real orders) and the code are withheld for proprietary reasons. Because both the synthetic generator and the transcode ground-truth labels are internal to the authors' pipeline, external readers cannot assess synthetic-to-real transfer or label correctness. Please provide a detailed description of the generator, the transcode language definition, the labeling/annotation process (including quality controls), and at least a redacted or synthetic public benchmark sample to enable reproducibility. Without this, the generalizability of the reported accuracy is not verifiable.
minor comments (5)
  1. [Table 1] Table 1 labels BART-base as 'Instruction-tuned', but BART-base (Lewis et al., 2019) is not an instruction-tuned model. Please correct the label or clarify what fine-tuning was performed on the model.
  2. [§5.3.1; throughout] The paper refers to Flan-T5-small and BART-base as 'LLMs'. These are encoder-decoder transformer models, not autoregressive large language models; more precise terminology would avoid confusion about the claimed comparison.
  3. [Table 3] The hyperparameter table has a formatting issue: the column 'accumulate_grad_batches1' appears misspelled and misaligned, and the row values 'N/A, 2, N/A, 4, 3' do not clearly map to the six model columns. Please fix the table layout.
  4. [§3.3] The phoneme transcription example for 'can I get a coffee please?' uses symbols such as '9', 'ě', and 'O' that are not standard IPA or ARPABET. Please clarify the phoneme inventory and transcription convention used by the ASR component.
  5. [§6] The final paragraph of §6 claims that the approach 'generalizes well across other spoken languages' due to its IPA-based representation, but no cross-lingual experiments are reported. Please temper or support this claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: experimental comparison is not self-referential; the unmeasured baseline-accuracy issue is an evidence gap, not a derivation loop.

full rationale

This is an empirical systems paper rather than a formal derivation, and I found no step where an output is defined in terms of the quantity it is supposed to predict, no fitted parameter that is later renamed as a prediction, and no load-bearing self-citation chain. The central comparison is between measured architectures on a held-out test set: Babylon, Atlantis, Camelot, Delphi, Flan-T5-small, and BART-base are all evaluated with the same transcode vocabulary and accuracy metric. The accuracy table does omit Flan-T5 and BART accuracy, with the footnote explaining that ASR errors 'significantly cascade down' to those models, but omitting a baseline measurement is a reporting and validity problem, not a circularity. Similarly, the paper selects the pooling stride k=4 for 'optimal performance' and the transcode representation is internally defined, but neither constitutes a prediction that reduces to its own construction: the model is still trained and evaluated on distinct inputs and ground-truth transcodes, and the claimed accuracy-latency-memory trade-off is an empirical claim whose main weakness is incompleteness, not self-reference. The proprietary synthetic training generator and the nondisclosed test set raise external-validity concerns, but the paper does not use the test set to define the metric or the model in a way that would make the evaluation circular. I therefore assign score 0, noting that the comparison against Flan-T5 and BART is currently unsupported by the reported numbers but that this is an evidentiary gap rather than a circularity.

Assumptions & free parameters 1 free parameters · 3 assumptions · 1 invented entities

The evaluation rests on the authors' internal transcode language, a proprietary synthetic-data generator, a tuned pooling stride, and a private test set; none of these have external anchors, so the reported accuracy cannot be independently verified.

free parameters (1)
  • token pooling stride k = 4
    Section 4.2 states 'we experimented with a range of k and selected k=4 for optimal performance.' No ablation or validation curve is shown; k directly affects input length, compute, and accuracy.
assumptions (3)
  • domain assumption The synthetic data generation method of Arel et al. (2019) produces training orders representative of real drive-thru speech, accents, dialects, and order patterns.
    Section 5.2 says the training set was generated with this method and evaluated on real drive-thru orders; the generator is not described and the data is withheld.
  • domain assumption Phoneme sequences produced by the upstream ASR are a sufficient input representation for NLU, and the test set contains realistic ASR errors.
    The system input is phonemes from ASR (Section 2); robustness claims rely on noisy test data, but no ASR error statistics are reported.
  • domain assumption Exact 1:1 match between predicted and ground-truth transcode sequences is a valid measure of customer intent correctness.
    Section 5.4 defines accuracy as exact match. The paper argues BLEU is not semantically informative, but it provides no human agreement or alternative semantic metric to validate the exact-match standard.
invented entities (1)
  • transcode regular language
    purpose: Encodes intents and slot fillers as a token sequence that the NLU model outputs, enabling multi-intent representation in one inference call.
    The transcode vocabulary and grammar are introduced for this system (Section 2) and are not tied to any external annotation scheme or public benchmark; accuracy is defined against the authors' own labels.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Transforming NLU with Babylon: A Case Study in Development of Real-time, Edge-Efficient, Multi-Intent Translation System for Automated Drive-Thru Ordering." pith.science (2026). https://pith.science/paper/4EBDBIPX

@misc{pith2026241115372,
  author       = {Pith},
  title        = {Pith review of: Transforming NLU with Babylon: A Case Study in Development of Real-time, Edge-Efficient, Multi-Intent Translation System for Automated Drive-Thru Ordering},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4EBDBIPX}},
  note         = {Machine review of arXiv:2411.15372}
}
read the original abstract

Real-time conversational AI agents face challenges in performing Natural Language Understanding (NLU) in dynamic, outdoor environments like automated drive-thru systems. These settings require NLU models to handle background noise, diverse accents, and multi-intent queries while operating under strict latency and memory constraints on edge devices. Additionally, robustness to errors from upstream Automatic Speech Recognition (ASR) is crucial, as ASR outputs in these environments are often noisy. We introduce Babylon, a transformer-based architecture that tackles NLU as an intent translation task, converting natural language inputs into sequences of regular language units ('transcodes') that encode both intents and slot information. This formulation allows Babylon to manage multi-intent scenarios in a single dialogue turn. Furthermore, Babylon incorporates an LSTM-based token pooling mechanism to preprocess phoneme sequences, reducing input length and optimizing for low-latency, low-memory edge deployment. This also helps mitigate inaccuracies in ASR outputs, enhancing system robustness. While this work focuses on drive-thru ordering, Babylon's design extends to similar noise-prone scenarios, for e.g. ticketing kiosks. Our experiments show that Babylon achieves significantly better accuracy-latency-memory footprint trade-offs over typically employed NMT models like Flan-T5 and BART, demonstrating its effectiveness for real-time NLU in edge deployment settings.

Figures

Figures reproduced from arXiv: 2411.15372 by the authors.

Figure 1
Figure 1. The input to our NLU model for this customer [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Illustration of our overall system. We re [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Architecture of Babylon. Blocks highlighted [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 15 canonical work pages

  1. [1]

    Armen Aghajanyan, Jean Maillard, Akshat Shrivastava, Keith Diedrick, Michael Haeger, Haoran Li, Yashar Mehdad, Veselin Stoyanov, Anuj Kumar, Mike Lewis, and Sonal Gupta. 2020. https://doi.org/10.18653/v1/2020.emnlp-main.408 Conversational semantic parsing . In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), ...

  2. [2]

    Itamar Arel, Joshua Benjamin Looks, Ali Ziaei, and Michael Lefkowitz. 2019. Conversational agent pipeline trained on synthetic data. US Patent 10,210,861

  3. [3]

    Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2014. Neural machine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473

  4. [4]

    Saptarashmi Bandyopadhyay and Tianyang Zhao. 2020. https://doi.org/10.18653/v1/2020.intexsempar-1.6 Natural language response generation from SQL with generalization and back-translation . In Proceedings of the First Workshop on Interactive and Executable Semantic Parsing, pages 46--49, Online. Association for Computational Linguistics

  5. [5]

    Peters, and Arman Cohan

    Iz Beltagy, Matthew E. Peters, and Arman Cohan. 2020. https://api.semanticscholar.org/CorpusID:215737171 Longformer: The long-document transformer . ArXiv, abs/2004.05150

  6. [6]

    Pu-Chin Chen, Henry Tsai, Srinadh Bhojanapalli, Hyung Won Chung, Yin-Wen Chang, and Chun-Sung Ferng. 2021. A simple and effective positional encoding for transformers. arXiv preprint arXiv:2104.08698

  7. [7]

    Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Yunxuan Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, et al. 2024. Scaling instruction-finetuned language models. Journal of Machine Learning Research, 25(70):1--53

  8. [8]

    Alexis Conneau, Kartikay Khandelwal, Naman Goyal, Vishrav Chaudhary, Guillaume Wenzek, Francisco Guzm \'a n, Edouard Grave, Myle Ott, Luke Zettlemoyer, and Veselin Stoyanov. 2020. https://doi.org/10.18653/v1/2020.acl-main.747 Unsupervised cross-lingual representation learning at scale . In Proceedings of the 58th Annual Meeting of the Association for Comp...

Show all 50 references
  1. [9]

    Le, and Ruslan Salakhutdinov

    Zihang Dai, Zhilin Yang, Yiming Yang, Jaime Carbonell, Quoc V. Le, and Ruslan Salakhutdinov. 2019. http://arxiv.org/abs/1901.02860 Transformer-xl: Attentive language models beyond a fixed-length context

  2. [10]

    Maxime De bruyn, Ehsan Lotfi, Jeska Buhmann, and Walter Daelemans. 2022. https://doi.org/10.18653/v1/2022.mmnlu-1.8 Machine translation for multilingual intent detection and slots filling . In Proceedings of the Massively Multilingual Natural Language Understanding Workshop (M...

  3. [11]

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. 2021. http://arxiv.org/abs/2010.11929 An image is worth 16x16 words:...

  4. [12]

    Razvan-Gabriel Dumitru, Darius Peteleaza, and Mihai Surdeanu. 2024. Enhancing transformer rnns with multiple temporal perspectives. arXiv preprint arXiv:2402.02625

  5. [13]

    Silvan Ferreira, Gustavo Leitão, Ivanovitch Silva, Allan Martins, and Paolo Ferrari. 2020. https://doi.org/10.1109/MetroInd4.0IoT48571.2020.9138181 Evaluating human-machine translation with attention mechanisms for industry 4.0 environment sql-based systems . In 2020 IEEE Inte...

  6. [14]

    Jack FitzGerald. 2020. https://aclanthology.org/2020.aacl-main.57 STIL - simultaneous slot filling, translation, intent classification, and language identification: Initial results using m BART on M ulti ATIS ++ . In Proceedings of the 1st Conference of the Asia-Pacific Chapte...

  7. [15]

    Matt Gardner, Pradeep Dasigi, Srinivasan Iyer, Alane Suhr, and Luke Zettlemoyer. 2018. https://doi.org/10.18653/v1/P18-5006 Neural semantic parsing . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics: Tutorial Abstracts, pages 17--18, M...

  8. [16]

    Alex Graves. 2012. http://arxiv.org/abs/1211.3711 Sequence transduction with recurrent neural networks

  9. [17]

    Gomez, and J¨urgen Schmidhuber

    Alex Graves, Santiago Fern´andez, Faustino J. Gomez, and J¨urgen Schmidhuber. https://api.semanticscholar.org/CorpusID:9901844 Connectionist temporal classification: Labelling unsegmented sequence data with recurrent neural networks

  10. [18]

    Daniel (Zhaohan) Guo, Gokhan Tur, Scott Wen-tau Yih, and Geoffrey Zweig. 2014. https://www.microsoft.com/en-us/research/publication/joint-semantic-utterance-classification-and-slot-filling-with-recursive-neural-networks-2/ Joint semantic utterance classification and slot filli...

  11. [19]

    Dilek Hakkani-Tür, Gokhan Tur, Asli Celikyilmaz, Yun-Nung Vivian Chen, Jianfeng Gao, Li Deng, and Ye-Yi Wang. 2016. https://www.microsoft.com/en-us/research/publication/multijoint/ Multi-domain joint semantic frame parsing using bi-directional rnn-lstm . In Proceedings of The ...

  12. [20]

    Adi Haviv, Ori Ram, Ofir Press, Peter Izsak, and Omer Levy. 2022. Transformer language models without positional encodings still learn positional information. arXiv preprint arXiv:2203.16634

  13. [21]

    Jakub Hoscilowicz, Pawel Pawlowski, Marcin Skorupa, Marcin Sowa \'n ski, and Artur Janicki. 2024. Large language models for expansion of spoken language understanding systems to new languages. arXiv preprint arXiv:2404.02588

  14. [22]

    Liang Huang, Senjie Liang, Feiyang Ye, and Nan Gao. 2024. https://doi.org/10.1109/TAI.2023.3309272 A fast attention network for joint intent detection and slot filling on edge devices . IEEE Transactions on Artificial Intelligence, 5(2):530--540

  15. [23]

    Yunpeng Huang, Jingwei Xu, Zixu Jiang, Junyu Lai, Zenan Li, Yuan Yao, Taolue Chen, Lijuan Yang, Zhou Xin, and Xiaoxing Ma. 2023. Advancing transformer architecture in long-context large language models: A comprehensive survey. arXiv preprint arXiv:2311.12351

  16. [24]

    Yue Ju, Alka Isac, and Yimin Nie. 2021. Chunkformer: Learning long time series with multi-stage chunked transformer. arXiv preprint arXiv:2112.15087

  17. [25]

    Gakuto Kurata, Bing Xiang, Bowen Zhou, and Mo Yu. 2016. https://doi.org/10.18653/v1/D16-1223 Leveraging sentence-level information with encoder LSTM for semantic slot filling . In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 207...

  18. [26]

    Mike Lewis, Yinhan Liu, Naman Goyal, Marjan Ghazvininejad, Abdelrahman Mohamed, Omer Levy, Veselin Stoyanov, and Luke Zettlemoyer. 2019. http://arxiv.org/abs/1910.13461 BART: denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehe...

  19. [27]

    Chin-Yew Lin. 2004. https://aclanthology.org/W04-1013 ROUGE : A package for automatic evaluation of summaries . In Text Summarization Branches Out, pages 74--81, Barcelona, Spain. Association for Computational Linguistics

  20. [28]

    Bing Liu and Ian Lane. 2016. Attention-based recurrent neural network models for joint intent detection and slot filling. arXiv preprint arXiv:1609.01454

  21. [29]

    Y Liu. 2020. Multilingual denoising pre-training for neural machine translation. arXiv preprint arXiv:2001.08210

  22. [30]

    Nicholas Locascio, Karthik Narasimhan, Eduardo DeLeon, Nate Kushman, and Regina Barzilay. 2016. https://api.semanticscholar.org/CorpusID:17360001 Neural generation of regular expressions from natural language with minimal domain knowledge . In Conference on Empirical Methods i...

  23. [31]

    Minh-Thang Luong, Hieu Pham, and Christopher D Manning. 2015. Effective approaches to attention-based neural machine translation. arXiv preprint arXiv:1508.04025

  24. [32]

    Dmitrii Marin, Jen-Hao Rick Chang, Anurag Ranjan, Anish Prabhu, Mohammad Rastegari, and Oncel Tuzel. 2021. http://arxiv.org/abs/2110.03860 Token pooling in vision transformers

  25. [33]

    Gr\' e goire Mesnil, Yann Dauphin, Kaisheng Yao, Yoshua Bengio, Li Deng, Dilek Hakkani-Tur, Xiaodong He, Larry Heck, Gokhan Tur, Dong Yu, and Geoffrey Zweig. 2015. Using recurrent neural networks for slot filling in spoken language understanding. IEEE/ACM Trans. Audio, Speech ...

  26. [34]

    Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. https://doi.org/10.3115/1073083.1073135 B leu: a method for automatic evaluation of machine translation . In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics, pages 311--3...

  27. [35]

    Libo Qin, Tianbao Xie, Wanxiang Che, and Ting Liu. 2021. https://api.semanticscholar.org/CorpusID:232110865 A survey on spoken language understanding: Recent advances and new frontiers . In International Joint Conference on Artificial Intelligence

  28. [36]

    Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J

    Colin Raffel, Noam M. Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, and Peter J. Liu. 2019. https://api.semanticscholar.org/CorpusID:204838007 Exploring the limits of transfer learning with a unified text-to-text transformer . J. Mach...

  29. [37]

    Tulika Saha, Neeti Priya, Sriparna Saha, and Pushpak Bhattacharyya. 2021. https://doi.org/10.1109/IJCNN52387.2021.9533525 A transformer based multi-task model for domain classification, intent detection and slot-filling . In 2021 International Joint Conference on Neural Networ...

  30. [38]

    Rahul Anand Sharma, Kunal Dhawan, and Balakrishna Pailla. 2021. https://api.semanticscholar.org/CorpusID:238226929 Phonetic word embeddings . ArXiv, abs/2109.14796

  31. [39]

    Marcin Sowański and Artur Janicki. 2023. https://doi.org/10.1109/IWSSIP58668.2023.10180229 Slot lost in translation? not anymore: A machine translation model for virtual assistants with type-independent slot transfer . In 2023 30th International Conference on Systems, Signals ...

  32. [40]

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, ukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in neural information processing systems, 30

  33. [41]

    Henry Weld, Xiaoqi Huang, Siqu Long, Josiah Poon, and Soyeon Caren Han. 2022. A survey of joint intent detection and slot filling models in natural language understanding. ACM Computing Surveys, 55(8):1--38

  34. [42]

    Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chaumond, Clement Delangue, Anthony Moi, Pierric 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, Mari...

  35. [43]

    Zhanghao Wu, Zhijian Liu, Ji Lin, Yujun Lin, and Song Han. 2020. Lite transformer with long-short range attention. arXiv preprint arXiv:2004.11886

  36. [44]

    Puyang Xu and Ruhi Sarikaya. 2013. https://doi.org/10.1109/ASRU.2013.6707709 Convolutional neural network based triangular crf for joint intent detection and slot filling . In 2013 IEEE Workshop on Automatic Speech Recognition and Understanding, pages 78--83

  37. [45]

    Puyang Xu and Ruhi Sarikaya. 2014. Targeted feature dropout for robust slot filling in natural language understanding. In INTERSPEECH, pages 258--262

  38. [46]

    Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, and Colin Raffel. 2021. https://doi.org/10.18653/v1/2021.naacl-main.41 m T 5: A massively multilingual pre-trained text-to-text transformer . In Proceedings of the 2021 Conferenc...

  39. [47]

    Wojciech Zaremba, Ilya Sutskever, and Oriol Vinyals. 2014. https://api.semanticscholar.org/CorpusID:17719760 Recurrent neural network regularization . ArXiv, abs/1409.2329

  40. [48]

    Ailing Zeng, Muxi Chen, Lei Zhang, and Qiang Xu. 2022. http://arxiv.org/abs/2205.13504 Are transformers effective for time series forecasting?

  41. [49]

    URL: " 'urlintro :=

    ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before...

  42. [50]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.