Pith. sign in

REVIEW 6 major objections 6 minor 1 cited by

Hybrid AI for Responsive Multi-Turn Online Conversations with Novel Dynamic Routing and Feedback Adaptation

T0 review · 6 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A hybrid chatbot router that answers easy queries with canned intents and routes hard ones to a retrieval pipeline claims 95% accuracy at 180 ms latency, beating either approach on its own.

desk verdict A sensible hybrid routing architecture whose headline numbers are not yet supported by the evaluation as written. read the letter →

arxiv 2506.02097 v2 pith:IHXUZMHV submitted 2025-06-02 cs.AI

Priyaranjan Pattnayak , Amit Agarwal , Hansa Meghwani , Hitesh Laxmichand Patel , Srikant Panda This is my paper
classification cs.AI
keywords hybridconversationalAIdynamicroutingintentclassificationretrieval-augmentedgenerationmulti-turndialoguefeedbackadaptationconfidencethresholdcannedresponses
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

The paper argues that enterprise chatbots do not have to choose between speed and accuracy. It proposes a system that classifies each incoming query by confidence: high-confidence queries get prewritten canned responses, low-confidence or ambiguous ones go to a retrieval-augmented generation (RAG) pipeline, and middle-confidence ones get a blended response. A dialogue context manager carries prior turns into classification, and a feedback loop adjusts the routing threshold and creates new intents from recurring unanswered queries. The reported result is 95% accuracy and 180 ms average latency on 10,000 in-house customer-support queries, with accuracy staying near 92% at 50,000 queries. The framework also reports better accuracy-latency trade-offs than either a canned-response baseline or a RAG pipeline alone.

What carries the argument

The load-bearing mechanism is dynamic confidence routing with context-aware thresholding. A fine-tuned BERT encoder turns a query into an embedding, the dialogue context manager augments that embedding with a sliding window of prior turns, and cosine similarity to predefined intent embeddings produces a confidence score $c$. Queries with $c > \tau_{FAQ}$ use canned responses, queries with $c \le \tau_{Out\text{-}of\text{-}Domain}$ go to RAG, and queries in between blend canned and RAG outputs through a language model. The feedback loop updates $\tau_{FAQ}$ as $\tau_{FAQ} = \tau_{FAQ} + \lambda \cdot (NFR - PFR)$, using positive and negative feedback rates, and creates new intents when clusters of unresolved queries pass a frequency threshold.

What would settle it

Have human annotators score the same 10,000 in-house system responses as correct or incorrect; if the human-correct rate is materially below 95%, the accuracy claim as stated fails. A cheaper check: vary the 90% cosine-similarity threshold from 0.85 to 0.95 on the same query set and see whether the proposed framework still beats the RAG pipeline on accuracy.

Watch

Extended reading notes

Core claim

The central claim is that a confidence-gated router can get the best of both modes: predefined intent responses for predictable queries and retrieval-augmented generation for everything else. The paper shows that this hybrid, combined with a dialogue context manager and a feedback-driven threshold update, reaches 95% accuracy and 180 ms average latency on an in-house evaluation set, compared with 53% accuracy at 68 ms for the canned-response baseline and 91% accuracy at 380 ms for the RAG pipeline alone. Category-level results attribute the gain to blending the two response types for mid-confidence queries, where the hybrid reaches 96% accuracy at 182 ms latency, and to preserving RAG-level accuracy on out-of-domain queries while cutting latency on FAQ queries by 82% relative to RAG.

Load-bearing premise

The headline accuracy number rests on defining a correct answer as one whose computer-measured semantic similarity to a prepared answer is at least 90%, with FAQ queries counted as perfectly correct when the right canned answer is retrieved; if that similarity measure does not match what users would call correct, the 95% figure is unsupported.

Editorial extensions

If this is right

  • FAQ-type queries can be answered at canned-response latency while complex queries retain RAG-level accuracy, so the system can serve both routine and edge-case customer requests from one endpoint.
  • The feedback loop gives a concrete rule for when to trust the canned-response path: negative feedback raises the confidence threshold, making the system more selective over time.
  • Multi-turn coherence is addressed directly: the dialogue context manager contributed to a turn efficiency of 1.7 turns per resolved query on the evaluation set, compared with 2.3 for RAG.
  • The scalability results imply that the routing design keeps accuracy above 92% even when the query load grows to 50,000, which matters for enterprise deployment.
  • The hybrid response formula for mid-confidence queries, $R_f = LLM(c \cdot R_c, (1-c) \cdot R_r)$, offers a simple way to combine deterministic and generative response sources.

Reading between the lines

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

  • The confidence-gating pattern could transfer to other high-volume NLP settings, such as code completion or search, where cheap pattern-matched outputs handle common cases and expensive generative models handle only the long tail; a testable extension is measuring how the optimal threshold drifts across domains.
  • The reported accuracy depends on the 90% cosine-similarity proxy for correctness, so a natural next study is comparing that metric against human judgments on the same response pairs to see whether the 95% number survives.
  • The feedback update is linear in positive and negative feedback rates, but the error analysis already shows that users sometimes down-rate correct responses because of slow response time; a more robust update would weight feedback by whether the complaint is about content or latency.
  • Because the paper does not isolate the dialogue context manager's contribution, a controlled ablation that removes context augmentation from Algorithm 1 would quantify how much of the multi-turn gain actually comes from that module.
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

6 major / 6 minor

Summary. This paper proposes a hybrid conversational AI framework that combines a fine-tuned BERT intent classifier with canned responses and a RAG pipeline. Queries are routed into FAQ (canned), contextual (hybrid), and out-of-domain (RAG) categories based on confidence thresholds, with a dialogue context manager for multi-turn coherence and a feedback loop that adjusts the FAQ threshold and creates new intents from clustered unresolved queries. The authors report 95% accuracy and 180ms average latency on an in-house 10,000-query OCI customer support dataset, claiming improvements over a canned-response baseline (53%, 68ms) and a RAG pipeline (91%, 380ms), plus scalability results up to 50,000 queries and a 500-sample human error analysis.

Significance. The proposed architecture is practically motivated and clearly described, and the inclusion of pseudo-code (Algorithm 1), a worked multi-turn example, and a human error analysis with inter-annotator agreement are strengths. If the headline numbers were supported by a calibrated evaluation, the framework would be a useful engineering contribution for enterprise customer-support chatbots, as it demonstrates a plausible way to trade off latency and accuracy. However, the paper's central evaluation claim is currently not established: the accuracy metric is partly definitional for FAQ queries, the cosine-similarity threshold is unvalidated, and the reported latency does not follow from the paper's own category-level table. The novelty over existing hybrid RAG-intent systems is also modest, since the core idea of routing by intent confidence has appeared in prior work; the main new elements are the feedback-driven threshold adaptation and context manager, whose individual contributions are not ablated.

major comments (6)
  1. [Section 4.1] The accuracy definition makes FAQ correctness equivalent to fetching the intended canned response (100% cosine similarity), and because high-confidence FAQ queries are routed to canned responses by construction, the FAQ accuracy component is a retrieval-success measure rather than an answer-quality measure. This inflates the proposed framework's accuracy relative to RAG, whose responses must satisfy a 90% cosine-similarity threshold against annotated answers. Report user-perceived correctness for FAQ responses (e.g., human judging of canned-response appropriateness) or apply the same semantic-similarity evaluation to all categories; otherwise the 95% headline is not interpretable.
  2. [Section 4.1] The 90% cosine-similarity threshold for contextual and out-of-domain responses is never calibrated against human judgments. The 500-sample manual error analysis in Section 4.3 and Table 8 is a separate qualitative exercise and does not validate the automatic metric. Provide a calibration study on a held-out sample showing how well the 90% threshold agrees with human binary correctness judgments, including precision and recall of the threshold choice, and report the distribution of similarity scores rather than only the pass rate.
  3. [Section 4.1 / Table 5 / Table 2] Section 4.1 fixes the evaluation mix at 40% FAQ, 30% contextual, and 30% out-of-domain, but Table 5 gives proposed-framework category latencies of 65ms, 182ms, and 379ms. The weighted mean is 0.4*65 + 0.3*182 + 0.3*379 = 194.3ms, not 180ms as reported in Tables 2 and 3. State the precise aggregation rule (e.g., whether follow-up turns or cache hits are excluded, or whether Table 5's latencies were measured under different conditions). Until this is resolved, the headline 180ms average latency is not supported by the paper's own disaggregated data.
  4. [Section 4.3 / Appendix D.1] Section 4.3 reports error-category percentages of 21%, 35%, 25%, and 19%, while Appendix D.1 reports 30%, 20%, and 10% for long multi-turn dialogues, retrieval inaccuracy, and feedback misalignment, with no reconciled figure for edge cases. These two accounts of the error analysis are mutually inconsistent. Align the numbers and state whether Appendix D.1 refers to the same 500-sample analysis or a different subset.
  5. [Section 4.1 / Tables 2-5] All accuracy and latency results are point estimates from a single evaluation set with no error bars, confidence intervals, or significance tests, and no code or data release is provided. Given that the paper's main claim is quantitative superiority over two baselines, report variance across multiple runs or bootstrap confidence intervals, and consider releasing the evaluation protocol and dataset descriptions to enable independent verification.
  6. [Section 3.2 / Section 4 / Section 6] The feedback-loop mechanism and the dialogue context manager are described as key contributions but are never evaluated in isolation. The experiments compare the full proposed system to baselines, not the system with and without the feedback loop or with and without the context manager. The paper's own Section 6 acknowledges that ablation studies are future work; consequently, the claims that feedback adaptation and multi-turn context management improve accuracy or latency are not supported by the reported experiments, and the novel contribution of dynamic routing cannot be separated from static-threshold routing.
minor comments (6)
  1. [Section 3.1 / Section 3.2] The initial thresholds are given as c > 0.85 for FAQ, 0.5 < c <= 0.85 for contextual, and c <= 0.5 for out-of-domain, while Section 3.2 uses tau_FAQ and tau_Out-of-Domain; clarify how the fixed defaults in Section 3.1 relate to the adaptive thresholds in Section 3.2 and Algorithm 1.
  2. [Section 3.2] The feedback update formula tau_FAQ = tau_FAQ + lambda * (NFR - PFR) uses rates rather than counts; specify how NFR and PFR are computed over the 100-interaction window and how the update keeps thresholds within [0,1].
  3. [Section 3.3 / Appendix A.6] Section 3.3 says storage uses OCI Opensearch, while Appendix A.6 says Elasticsearch; reconcile these conflicting descriptions of the prototype.
  4. [Appendix C] Appendix C references Figure 2 for the response-blending prompt, but the figure content is not present in the text; either include the figure or provide the prompt inline.
  5. [References] Several references are incomplete or malformed, including 'Pattnayak et al.' with no year, 'V olume3' in the Patel et al. 2025 entry, and the reversed 'et al. Zhao, W. 2020' entry; a full reference cleanup is needed.
  6. [Section 4.3] Section 4.3 says '32 (6%) samples were found erroneous'; 32/500 is 6.4%, so the parenthetical should be corrected to 6.4% or the counts should be reconciled.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation identified; the headline accuracy and latency numbers are empirical measurements rather than results forced by construction, though internal arithmetic inconsistencies and an uncalibrated similarity metric remain non-circular validity concerns.

full rationale

I found no circular step that reduces a claimed result to its own inputs by construction. The framework routes queries by cosine-similarity confidence thresholds and the evaluation measures accuracy and latency on an in-house dataset; the same accuracy rule is applied to the canned-response and RAG baselines, so the comparison is not definitionally rigged. For FAQ queries, fetching the intended canned response is scored as 100% cosine similarity, but this is a task definition rather than a circular derivation because the system can still fetch the wrong response and the baselines are scored under the same rule. The feedback-loop and dialogue-context-manager contributions are asserted without ablation studies, which is missing support rather than circularity. Several self-citations appear (e.g., Pattnayak et al., 2024, for the sub-200ms latency standard), but none is load-bearing for the central experimental claims. Reproducibility is weakened by the mismatch between the stated 40/30/30 category mix and Table 5 latencies (0.4*65 + 0.3*182 + 0.3*379 = 194.3ms, not the reported 180ms) and by inconsistent error distributions between Section 4.3 and Appendix D.1; these are reporting and validity flaws, not evidence of circular reasoning.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central claims depend on several unverified domain assumptions and hand-chosen thresholds. No external benchmark is used, and the evaluation data and code are private.

free parameters (5)
  • FAQ confidence threshold tau_FAQ = 0.85 (default, adjusted online via feedback)
    Determines whether a query is answered by a canned response or routed to RAG; hand-set default with no principled derivation.
  • Out-of-domain threshold tau_OOD = 0.5 (kept constant)
    Defines the out-of-domain classification boundary; chosen ad hoc and explicitly held constant.
  • Feedback scaling factor lambda = not specified
    Controls how strongly feedback changes tau_FAQ; the value is not reported, yet the feedback adaptation claims depend on it.
  • Accuracy similarity threshold = 90% cosine similarity
    Defines what counts as a correct contextual or out-of-domain response; directly determines reported accuracy with no justification.
  • Dataset category proportions = 40% FAQ, 30% contextual, 30% OOD, 20% multi-turn follow-ups
    Chosen evaluation mix that likely favors the hybrid router; no sensitivity analysis over category proportions is provided.
assumptions (5)
  • domain assumption Cosine similarity between a generated response and an annotated ground-truth answer is a valid measure of response accuracy.
    Section 4.1 defines accuracy using this similarity; no correlation with human-judged correctness is provided.
  • domain assumption The in-house OCI customer-support dataset and its ground-truth annotations are representative of real enterprise queries.
    Sections 4.1 and A.3 describe 10,000 queries from internal data; the dataset is not released, so representativeness cannot be checked.
  • domain assumption User thumbs-up and thumbs-down feedback reliably signals response quality and can be used to adjust thresholds.
    Section 3.2 uses feedback to update tau_FAQ while Appendix A.5 notes 19% feedback misinterpretation, weakening this premise.
  • domain assumption BERT embeddings and the in-house bi-encoder and cross-encoder capture the semantic similarity needed for intent confidence.
    Algorithm 1 and Section 3.1 rely on embedding cosine similarity for confidence, but no calibration or confidence evaluation is provided.
  • domain assumption The RAG baseline is representative of state-of-the-art RAG systems.
    Section 4.2 compares against a single RAG pipeline with no details on retrieval or generation configuration, so the baseline may be understated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hybrid AI for Responsive Multi-Turn Online Conversations with Novel Dynamic Routing and Feedback Adaptation." pith.science (2026). https://pith.science/paper/IHXUZMHV

@misc{pith2026250602097,
  author       = {Pith},
  title        = {Pith review of: Hybrid AI for Responsive Multi-Turn Online Conversations with Novel Dynamic Routing and Feedback Adaptation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IHXUZMHV}},
  note         = {Machine review of arXiv:2506.02097}
}
read the original abstract

Retrieval-Augmented Generation (RAG) systems and large language model (LLM)-powered chatbots have significantly advanced conversational AI by combining generative capabilities with external knowledge retrieval. Despite their success, enterprise-scale deployments face critical challenges, including diverse user queries, high latency, hallucinations, and difficulty integrating frequently updated domain-specific knowledge. This paper introduces a novel hybrid framework that integrates RAG with intent-based canned responses, leveraging predefined high-confidence responses for efficiency while dynamically routing complex or ambiguous queries to the RAG pipeline. Our framework employs a dialogue context manager to ensure coherence in multi-turn interactions and incorporates a feedback loop to refine intents, dynamically adjust confidence thresholds, and expand response coverage over time. Experimental results demonstrate that the proposed framework achieves a balance of high accuracy (95\%) and low latency (180ms), outperforming RAG and intent-based systems across diverse query types, positioning it as a scalable and adaptive solution for enterprise conversational AI applications.

Figures

Figures reproduced from arXiv: 2506.02097 by the authors.

Figure 1
Figure 1. High-level Architecture of the Hybrid Framework. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Prompt for Blending Responses D Feature Limitations & Related Future Work D.1 Limitations Despite the strong performance of the proposed framework on a variety of metrics, certain feature￾specific limitations remain: • Edge Cases in Intent Classification: Ambigu￾ous queries near confidence thresholds may cause inconsistencies, as seen in our error analysis, where subtle semantic differences led to misclassification.… view at source ↗

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Generative Voice Bursts during Phone Call

    cs.SD 2025-06 reject novelty 5.0 of 10

    A proposed system that would generate and deliver 3 to 5 second AI voice bursts during an active phone call to convey emergency information, but it is not backed by any implementation or test.

Reference graph

Works this paper leans on

41 extracted references · 26 canonical work pages · cited by 1 Pith paper

  1. [1]

    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.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...

  2. [2]

    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 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Amit Agarwal, Srikant Panda, Angeline Charles, Bhargava Kumar, Hitesh Patel, Priyaranjan Pattnayak, Taki Hasan Rafi, Tejaswini Kumar, and Dong-Kyu Chae. 2025 a . http://arxiv.org/abs/2412.19794 Mvtamperbench: Evaluating robustness of vision-language models

    work page Pith review arXiv 2025

  4. [4]

    Amit Agarwal, Srikant Panda, and Kulbhushan Pachauri. 2025 b . https://aclanthology.org/2025.coling-industry.9/ Fs-dag: Few shot domain adapting graph networks for visually rich document understanding . In Proceedings of the 31st International Conference on Computational Linguistics: Industry Track, pages 100--114, Abu Dhabi, UAE. Association for Computat...

    work page 2025

  5. [5]

    Amit Agarwal, Hitesh Patel, Priyaranjan Pattnayak, Srikant Panda, Bhargava Kumar, and Tejaswini Kumar. 2024. Enhancing document ai data generation through graph-based synthetic layouts. arXiv preprint arXiv:2412.03590

    arXiv 2024

  6. [6]

    Jason Bordes et al. 2020. https://arxiv.org/abs/2004.05344 Contextualized end-to-end learning for conversational ai . In Proceedings of the 2020 Conference on Neural Information Processing Systems (NeurIPS)

    work page Pith review arXiv 2020

  7. [7]

    Florin Cuconasu, Giovanni Trappolini, Federico Siciliano, Simone Filice, Cesare Campagnano, Yoelle Maarek, Nicola Tonellotto, and Fabrizio Silvestri. 2024. https://doi.org/10.1145/3626772.3657834 The power of noise: Redefining retrieval for rag systems . In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Informati...

    arXiv 2024

  8. [8]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, et al. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. Association for Computational Linguistics (ACL), pages 4171--4186

    work page 2019

Show all 41 references
  1. [9]

    Matthijs Douze, Alexandr Guzhva, Chengqi Deng, Jeff Johnson, Gergely Szilvasy, Pierre-Emmanuel Mazaré, Maria Lomeli, Lucas Hosseini, and Hervé Jégou. 2024. http://arxiv.org/abs/2401.08281 The faiss library

    2024 arXiv

  2. [10]

    Friedrich et al

    A. Friedrich et al. 2020. https://arxiv.org/abs/2211.16175 Context-aware robust fine-tuning for chatbots . In Proceedings of the 2020 International Conference on AI and Machine Learning

    2020 arXiv

  3. [11]

    Hao Gao, Dongxu Li, Shuohang Wang, and Wenjie Li. 2020 a . Hybrid conversational frameworks for multi-domain dialogue systems. In ACL, pages 298--305

    2020

  4. [12]

    Hao Gao, Dongxu Li, Liheng Xu, Shuohang Wang, and Wenjie Li. 2020 b . https://arxiv.org/abs/2003.05499 Search-augmented generation for dialogue systems . In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL)

    2020

  5. [13]

    Pengfei Gao, Jianfei Gao, and et al. 2020 c . Modular graph networks for reasoning over text. ACL

    2020

  6. [14]

    Miguel Grinberg. 2018. Flask web development: Developing web applications with Python. O'Reilly Media

    2018

  7. [15]

    Ricardo Hernandez, Rahul Gupta, and Shubham Patel. 2020. https://arxiv.org/abs/2001.05994 Efficient and scalable hybrid retrieval for search engines . In Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval

    2020 arXiv

  8. [16]

    Paulina Toro Isaza, Michael Nidd, Noah Zheutlin, Jae wook Ahn, Chidansh Amitkumar Bhatt, Yu Deng, Ruchi Mahindru, Martin Franz, Hans Florian, and Salim Roukos. 2024. http://arxiv.org/abs/2409.13707 Retrieval augmented generation-based incident resolution recommendation system ...

    2024 arXiv

  9. [17]

    Vladimir Karpukhin, Barlas Oguz, Sewon Min, et al. 2020. Dense passage retrieval for open-domain question answering. Empirical Methods in Natural Language Processing (EMNLP), pages 6769--6781

    2020

  10. [18]

    Mandar Kulkarni, Praveen Tangarajan, Kyung Kim, and Anusua Trivedi. 2024. http://arxiv.org/abs/2401.06800 Reinforcement learning for optimizing rag for domain chatbots

    2024 arXiv

  11. [19]

    Jinhyuk Lee et al. 2020. https://arxiv.org/abs/2407.08223 Speculative rag: Enhancing retrieval augmented generation through drafting . In Proceedings of the 2020 Conference on Neural Information Processing Systems (NeurIPS)

    2020 arXiv

  12. [20]

    Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vassilios Stamatescu, Tim Rockt \"a schel, Sebastian Ruder, Pontus Stenetorp, and LUKAS RICHTER. 2020. https://arxiv.org/abs/2005.11401 Retrieval-augmented generation for knowledge-intensive nlp tasks . In Proceedin...

    2020 arXiv

  13. [21]

    Hansa Meghwani, Amit Agarwal, Priyaranjan Pattnayak, Hitesh Laxmichand Patel, and Srikant Panda. 2025. http://arxiv.org/abs/2505.18366 Hard negative mining for domain-specific retrieval in enterprise systems

    2025 arXiv

  14. [22]

    Michelson et al

    J. Michelson et al. 2020. https://arxiv.org/abs/2410.12444 Expanding chatbot knowledge in customer service: Context-aware similar question generation using large language models . In Proceedings of the 2020 Conference on Natural Language Processing

    2020

  15. [23]

    Bingqing Mitra, Karan Goel, and Nikita Soni. 2021. https://arxiv.org/abs/2105.05841 Hybrid retrieval and ranking methods for information retrieval . In Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval

    2021 arXiv

  16. [24]

    Hitesh Laxmichand Patel, Amit Agarwal, Arion Das, Bhargava Kumar, Srikant Panda, Priyaranjan Pattnayak, Taki Hasan Rafi, Tejaswini Kumar, and Dong-Kyu Chae. 2025. Sweeval: Do llms really swear? a safety benchmark for testing limits for enterprise use. In Proceedings of the 202...

    2025

  17. [25]

    Hitesh Laxmichand Patel, Amit Agarwal, Bhargava Kumar, Karan Gupta, and Priyaranjan Pattnayak. 2024. Llm for barcodes: Generating diverse synthetic data for identity documents. arXiv preprint arXiv:2411.14962

    2024 arXiv

  18. [26]

    Priyaranjan Pattnayak. 2017. Predicting rainfall at seattle tacoma airport-logistic regression

    2017

  19. [27]

    Review of reference generation methods in large language models

    Priyaranjan Pattnayak, Amit Agarwal, Bhargava Kumar, Yeshil Bangera, Srikant Panda, Tejaswini Kumar, and Hitesh Laxmichand Patel. Review of reference generation methods in large language models. Journal ID, 9339:1263

  20. [28]

    Priyaranjan Pattnayak, Hitesh Laxmichand Patel, and Amit Agarwal. 2025 a . http://arxiv.org/abs/2504.16977 Tokenization matters: Improving zero-shot ner for indic languages

    2025 arXiv

  21. [30]

    Priyaranjan Pattnayak, Hitesh Laxmichand Patel, Amit Agarwal, Bhargava Kumar, Srikant Panda, and Tejaswini Kumar. 2025 c . http://arxiv.org/abs/2502.13108 Improving clinical question answering with multi-task learning: A joint approach for answer extraction and medical categorization

    2025 arXiv

  22. [31]

    Priyaranjan Pattnayak, Hitesh Laxmichand Patel, Bhargava Kumar, Amit Agarwal, Ishan Banerjee, Srikant Panda, and Tejaswini Kumar. 2024. Survey of large multimodal model datasets, application categories and taxonomy. arXiv preprint arXiv:2412.17759

    2024 arXiv

  23. [32]

    Nils Reimers and Iryna Gurevych. 2019. Sentence-bert: Sentence embeddings using siamese bert-networks. arXiv preprint arXiv:1908.10084

    2019 arXiv

  24. [33]

    S. E. Robertson and S. Walker. 1994. https://doi.org/10.1007/978-1-4471-2099-5_24 Some Simple Effective Approximations to the 2-Poisson Model for Probabilistic Weighted Retrieval , pages 232--241. Springer London

    1994 doi

  25. [34]

    Tim Rockt \"a schel, Sebastian Ruder, Shinnosuke Takamatsu, and Pontus Stenetorp. 2020. https://arxiv.org/abs/2004.08789 Rethinking the role of knowledge in dialogue systems . In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL)

    2020 arXiv

  26. [35]

    Victor Sanh, Thomas Wolf, Julien Chaumond, and Cl \'e ment Delangue. 2020 a . Multitask mixture of sequence generation tasks for diverse natural language generation. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 1962--1971

    2020

  27. [36]

    Victor Sanh, Thomas Wolf, Julien Chaumond, Clement Delangue, Pierrick Sprechmann, Alex Wang, Shinnosuke Takamatsu, and Tim Rockt \"a schel. 2020 b . https://arxiv.org/abs/2002.08909 Realm: Retrieval-augmented language model pre-training . In Proceedings of the 37th Internation...

    2020 arXiv

  28. [37]

    Serban, Alessandro Sordoni, Yoshua Bengio, and Jian-Yun Nie

    Iulian V. Serban, Alessandro Sordoni, Yoshua Bengio, and Jian-Yun Nie. 2017. https://arxiv.org/abs/1706.05039 A survey of available corpora for building data-driven dialogue systems . In Proceedings of the 2nd Workshop on Dialogue Systems Technology Evaluation (DST'17)

    2017 arXiv

  29. [38]

    Vishal Shah, Pushpak Bhattacharyya, and Khurshid Ahmad. 2018. https://arxiv.org/abs/1807.07793 Building end-to-end dialogue systems with transformer models . In Proceedings of the 22nd Conference on Computational Natural Language Learning (CoNLL)

    2018 arXiv

  30. [39]

    Kurt Shuster, Eric Smith, Stephen Roller, Y-Lan Boureau, and Jason Weston. 2021. Retrieval-augmented generation for knowledge-intensive nlp tasks. arXiv preprint arXiv:2102.09527

    2021 arXiv

  31. [40]

    Yi Tay, Mostafa Dehghani, Samira Abnar, Dara Bahri, Yikang Shen, Xingdi Zhou, and Donald Metzler. 2023. Efficient and scalable nlp with small and large pretrained language models. arXiv preprint arXiv:2305.13249

    2023 arXiv

  32. [41]

    Sriram Veturi, Saurabh Vaichal, Reshma Lal Jagadheesh, Nafis Irtiza Tripto, and Nian Yan. 2024. http://arxiv.org/abs/2409.03708 Rag based question-answering for contextual response prediction system

    2024 arXiv

  33. [42]

    et al. Zhao, W. 2020. https://arxiv.org/abs/2002.08735 A retrieval-augmented encoder-decoder for knowledge-intensive nlp tasks . In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (ACL)

    2020 arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.