Pith. sign in

REVIEW 4 major objections 4 minor 1 cited by

From Ranking to Selection: A Simple but Efficient Dynamic Passage Selector for Retrieval Augmented Generation

T0 review · 4 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read A dynamic passage selector that adapts evidence size outperforms fixed Top-K reranking in retrieval-augmented generation.

desk verdict The abstract claims a 30% F1 jump on MuSiQue but withholds all experimental detail, so the paper is unverifiable from what we have; if the full text backs up the numbers, it's worth refereeing. read the letter →

arxiv 2508.09497 v1 pith:E5D6KLYI submitted 2025-08-13 cs.CL cs.AI

Siyuan Meng , Junming Liu , Yirong Chen , Song Mao , Pinlong Cai , Guohang Yan , Botian Shi , Ding Wang This is my paper
classification cs.CLcs.AI
keywords Retrieval-AugmentedGenerationRerankingDynamicPassageSelectionMulti-hopQuestionAnsweringInter-passageDependenciesSupervisedLearning
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

Retrieval-augmented generation usually retrieves a fixed number of passages, and rerankers score each passage alone before cutting at a predetermined Top-K. This paper claims that this fixed-size selection hurts multi-hop questions, where evidence must be combined across documents, and introduces Dynamic Passage Selector (DPS), which treats passage selection as a supervised learning problem and captures dependencies between candidate passages. DPS is designed as a plug-and-play reranking module that adaptively selects the set of passages to feed the generator. The authors report that DPS outperforms strong rerankers and fine-tuning baselines on five benchmarks, including F1 gains of 30.06% and 15.4% over two leading rerankers on MuSiQue.

What carries the argument

The Dynamic Passage Selector (DPS) — a fine-tuned reranking module that models passage selection as supervised prediction over a set of candidates. Instead of assigning each passage an independent relevance score and choosing a fixed K, DPS captures inter-passage dependencies and outputs a variable-size set of passages for the generator. It is designed to plug into an existing RAG pipeline without altering retrieval or generation.

What would settle it

A direct test: fix a sweep of Top-K values for each baseline reranker on MuSiQue and compare the best fixed-K F1 to DPS. If a well-tuned fixed K matches or beats DPS, the core claim of dynamic selection being superior collapses.

Watch

Extended reading notes

Core claim

The central discovery is that reranking quality can be improved by replacing independent passage scoring plus a fixed Top-K cutoff with a model fine-tuned to output a subset of passages that jointly support the answer. DPS is that model: it encodes the candidate set, learns inter-passage dependencies, and predicts a dynamic selection. The paper shows consistent superiority over point-wise and list-wise rerankers and fine-tuning methods, with the largest wins on the multi-hop MuSiQue dataset, where F1 rises by 30.06% over one strong reranker and 15.4% over another. The implication is that adaptive, dependency-aware evidence selection materially strengthens reasoning in complex RAG scenarios.

Load-bearing premise

The gains rest on the availability of supervised labels for good passage sets and on the learned inter-passage dependencies generalizing to unseen multi-hop queries.

Editorial extensions

If this is right

  • If DPS is correct, RAG systems should stop treating K as a universal constant and let the selector adapt the evidence budget to the query.
  • DPS's plug-and-play design means retrieval and generation components can be left unchanged while the reranking stage is upgraded.
  • The MuSiQue results indicate that dependency-aware selection is especially valuable when an answer must be synthesized from multiple documents.
  • Treating selection as supervised learning opens a path to training rerankers on downstream task success rather than on relevance labels alone.

Reading between the lines

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

  • An untested consequence is that dynamic selection might also help single-hop QA by trimming noisy passages; the paper's benchmarks are primarily multi-hop.
  • The relative gain over fixed-K baselines depends on how each baseline's K was tuned; sweeping K for all methods would test whether the advantage is intrinsic to adaptive selection.
  • A natural follow-up is to use generator confidence as a training signal for the selector, closing the loop between selection and answer quality.
  • Since DPS is fine-tuned, its transfer across domains will depend on the diversity of the training set; the abstract does not report zero-shot generalization.
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 / 4 minor

Summary. The paper proposes Dynamic Passage Selector (DPS), a reranking framework that reformulates passage selection as a supervised learning problem, aiming to capture inter-passage dependencies and dynamically select a relevant set of passages for RAG. The abstract claims consistent improvements over state-of-the-art rerankers and fine-tuning methods across five benchmarks, with a headline 30.06% F1 improvement over Qwen3-reranker on MuSiQue. The manuscript as provided contains only the abstract; no experimental protocol, model architecture, training objective, or quantitative comparisons are described.

Significance. If the empirical claims are substantiated, the contribution is significant: dynamic passage selection that accounts for inter-passage dependencies could address a known limitation of fixed Top-K reranking in multi-hop RAG. The conceptual direction is timely and plausible. However, the current manuscript provides no evidence beyond the abstract, and the magnitude of the reported gains (e.g., 30.06% on MuSiQue) is implausibly large without a controlled evaluation. No code, data, or reproducible experimental details are supplied, so the significance cannot be assessed. The contribution is a promising hypothesis, not yet a verified result.

major comments (4)
  1. [Abstract] The central empirical claim—'DPS improves the F1-score by 30.06% and 15.4% over strong baselines' on MuSiQue—is unsupported by any experimental protocol. The abstract does not specify the retriever, candidate pool size, passage/reranker input length, baseline fine-tuning status, selection of K for fixed-K baselines, evaluation splits, or statistical significance. A gain of this magnitude could arise from an uneven configuration (e.g., DPS seeing a larger candidate set or choosing K adaptively while baselines are assigned a poorly chosen fixed K). This is a load-bearing omission, not a presentation issue.
  2. [Abstract] The method is described as 'fine-tuned to capture inter-passage dependencies and dynamically select the most relevant set of passages,' but the training supervision is unspecified. How are the 'most relevant set' labels obtained? If they are derived from gold supporting facts, downstream answer correctness, or any query-dependent signal, the learned dependencies may fit dataset artifacts and fail to generalize. Without a precise description of label construction and train/test separation, the risk of leakage or overfitting cannot be evaluated.
  3. [Manuscript structure] The submitted manuscript consists solely of the abstract. There is no method section, no equations, no pseudocode, and no experimental section. The central claim depends on a trained model whose architecture, loss function, and optimization are never described. This prevents any assessment of the proposed mechanism ('inter-passage dependencies') and makes the work non-reproducible. The missing material is not a minor stylistic point; it is the evidence for the claimed advantage.
  4. [Abstract] The comparison against 'state-of-the-art rerankers' is potentially confounded by the dynamic selection itself. If DPS selects a variable number of passages while baselines are constrained to a fixed K, the improvement may reflect the benefit of adaptive K rather than the learned inter-passage scoring. A controlled experiment should compare DPS against fixed-K baselines using an equivalent expected K, or include an oracle-K baseline, to isolate the contribution of the learned selection policy.
minor comments (4)
  1. [Abstract] The phrase 'fine-tuning methods' is vague; it should name the specific methods compared (e.g., RankingGPT, Qwen3-reranker) and clarify whether these are used as rerankers or as fine-tuning approaches.
  2. [Abstract] The number '30.06%' suggests a single run with no variance. Reporting a point estimate without error bars, confidence intervals, or multiple seeds makes the result difficult to interpret, especially given the small number of benchmarks.
  3. [Abstract] The claim 'seamless plug-and-play module' is not supported by any system integration results; the abstract does not describe how DPS interfaces with different retrievers or generators.
  4. [Abstract] No references to prior work are provided, so the novelty cannot be checked. The abstract should at least cite representative reranking approaches and recent RAG systems to contextualize the contribution.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the abstract reports an empirical supervised-learning method evaluated on held-out benchmarks, with no derivation chain that reduces to its inputs.

full rationale

The provided manuscript contains only the abstract; no equations, derivations, or self-citations are present. The central claim is a performance comparison on five benchmarks, which is an empirical claim about held-out generalization. The method is described as a supervised learning system that is fine-tuned to select passage sets, but this is a standard train/evaluate setup: the training supervision is an input to the model, not a quantity being re-derived or renamed as a prediction. No fitted parameter is called a prediction, no uniqueness theorem is imported from the authors, and no ansatz is smuggled via citation. The abstract gives insufficient experimental detail to verify the reported gains, but that is a reproducibility/correctness concern, not circularity. Therefore the circularity score is 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The paper introduces no new entities, only a learned reranking module. The main unstated premises are about the learnability and transferability of inter-passage dependencies, and about the validity of the evaluation setup. Since only the abstract was available, this ledger is based on the abstract's explicit statements.

assumptions (3)
  • domain assumption Inter-passage dependencies can be captured by fine-tuning a reranker on passage selection labels.
    Stated in abstract: 'DPS is fine-tuned to capture inter-passage dependencies.' This is the core modeling assumption that motivates the method.
  • domain assumption Supervised learning on passage selection generalizes to unseen multi-hop queries.
    The claim of consistent outperformance assumes the trained selector generalizes beyond its training data. This is not proven in the abstract.
  • domain assumption The five evaluation benchmarks are representative of complex RAG scenarios and are fairly compared.
    The abstract states 'comprehensive evaluations on five benchmarks' without details on test conditions or baseline configuration. The claimed improvements depend on the fairness of these comparisons.

how reviews work

0 comments
Cite this review

Pith. "Pith review of From Ranking to Selection: A Simple but Efficient Dynamic Passage Selector for Retrieval Augmented Generation." pith.science (2026). https://pith.science/paper/E5D6KLYI

@misc{pith2026250809497,
  author       = {Pith},
  title        = {Pith review of: From Ranking to Selection: A Simple but Efficient Dynamic Passage Selector for Retrieval Augmented Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/E5D6KLYI}},
  note         = {Machine review of arXiv:2508.09497}
}
read the original abstract

Retrieval-augmented generation (RAG) systems are often bottlenecked by their reranking modules, which typically score passages independently and select a fixed Top-K size. This approach struggles with complex multi-hop queries that require synthesizing evidence across multiple documents, creating a trade-off where small K values omit crucial information and large K values introduce noise. To address this, we introduce the Dynamic Passage Selector (DPS), a novel reranking framework that treats passage selection as a supervised learning problem. Unlike traditional point-wise or list-wise methods, DPS is fine-tuned to capture inter-passage dependencies and dynamically select the most relevant set of passages for generation. As a seamless plug-and-play module, DPS requires no modifications to the standard RAG pipeline. Comprehensive evaluations on five benchmarks show that DPS consistently outperforms state-of-the-art rerankers and fine-tuning methods. Notably, on the challenging MuSiQue dataset, DPS improves the F1-score by 30.06% and 15.4% over strong baselines like Qwen3-reranker and RankingGPT, respectively. Our results demonstrate that by enabling adaptive evidence selection, DPS substantially enhances reasoning capabilities in complex RAG scenarios.

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. MemVerse: Multimodal Memory for Lifelong Learning Agents

    cs.AI 2025-12 reject novelty 4.0 of 10

    MemVerse reports large gains on multimodal benchmarks by adding a hierarchical knowledge-graph memory plus fine-tuned parametric recall, but its strongest video-retrieval result uses ground-truth caption-video pairs i...

Reference graph

Works this paper leans on

45 extracted references · 32 canonical work pages · cited by 1 Pith paper

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  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]

    F.; Craswell, N.; Deng, L.; Gao, J.; Liu, X.; Majumder, R.; McNamara, A.; Mitra, B.; Nguyen, T

    Bajaj, P.; Campos, D. F.; Craswell, N.; Deng, L.; Gao, J.; Liu, X.; Majumder, R.; McNamara, A.; Mitra, B.; Nguyen, T. M.; Rosenberg, M.; Song, X.; Stoica, A. M.; Tiwary, S.; and Wang, T. 2016. MS MARCO: A Human Generated MAchine Reading COmprehension Dataset. arXiv: Computation and Language

    work page 2016

  4. [4]

    Chen, J.; Xiao, S.; Zhang, P.; Luo, K.; Lian, D.; and Liu, Z. 2024 a . Bge m3-embedding: Multi-lingual, multi-functionality, multi-granularity text embeddings through self-knowledge distillation. arXiv preprint arXiv:2402.03216

    arXiv 2024

  5. [5]

    Chen, J.; Xiao, S.; Zhang, P.; Luo, K.; Lian, D.; and Liu, Z. 2024 b . BGE M3-Embedding: Multi-Lingual, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation. In Annual Meeting of the Association for Computational Linguistics

    work page 2024

  6. [6]

    Cohere Inc. 2024. Rerank. https://cohere.com/rerank. Accessed: 2024-11-24

    work page 2024

  7. [7]

    Dong, G.; Zhu, Y.; Zhang, C.; Wang, Z.; Dou, Z.; and Wen, J.-R. 2024. Understand What LLM Needs: Dual Preference Alignment for Retrieval-Augmented Generation. Proceedings of the ACM on Web Conference 2025

    work page 2024

  8. [8]

    Edge, D.; Trinh, H.; Cheng, N.; Bradley, J.; Chao, A.; Mody, A.; Truitt, S.; and Larson, J. 2024. From Local to Global: A Graph RAG Approach to Query-Focused Summarization. ArXiv, abs/2404.16130

    arXiv 2024

Show all 45 references
  1. [9]

    Gao, J.; Chen, B.; Zhao, X.; Liu, W.; Li, X.; Wang, Y.; Wang, W.; Guo, H.; and Tang, R. 2025. LLM4Rerank: LLM-based Auto-Reranking Framework for Recommendations. In Proceedings of the ACM on Web Conference 2025, WWW '25, 228–239. New York, NY, USA: Association for Computing Ma...

    2025

  2. [10]

    A.; and Proença, H

    Hambarde, K. A.; and Proença, H. 2023. Information Retrieval: Recent Advances and Beyond. IEEE Access, 11: 76581--76604

    2023

  3. [11]

    D.; Sugawara, S.; and Aizawa, A

    Ho, X.; Nguyen, A.-K. D.; Sugawara, S.; and Aizawa, A. 2020. Constructing a multi-hop qa dataset for comprehensive evaluation of reasoning steps. arXiv preprint arXiv:2011.01060

    2020 arXiv

  4. [12]

    Kang, J.; Li, R.; Liu, Q.; Huang, Z.; Zhang, Z.; Chen, Y.; Zhu, L.; and Su, Y. 2025. Distribution-Driven Dense Retrieval: Modeling Many-to-One Query-Document Relationship. Proceedings of the AAAI Conference on Artificial Intelligence, 39(11): 11933--11941

    2025

  5. [13]

    u ttler, H.; Lewis, M.; Yih, W.-t.; Rockt\

    Lewis, P.; Perez, E.; Piktus, A.; Petroni, F.; Karpukhin, V.; Goyal, N.; K\" u ttler, H.; Lewis, M.; Yih, W.-t.; Rockt\" a schel, T.; Riedel, S.; and Kiela, D. 2020. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. In Larochelle, H.; Ranzato, M.; Hadsell, R.; ...

    2020

  6. [14]

    Li, X.; Zhu, Y.; Liu, S.; Ju, J.; Qu, Y.; and Cheng, G. 2023. DyRRen: A Dynamic Retriever-Reranker-Generator Model for Numerical Reasoning over Tabular and Textual Data. Proceedings of the AAAI Conference on Artificial Intelligence, 37(11): 13139--13147

    2023

  7. [15]

    Li, Y.; Yang, N.; Wang, L.; Wei, F.; and Li, W. 2024. Learning to Rank in Generative Retrieval. Proceedings of the AAAI Conference on Artificial Intelligence, 38(8): 8716--8723

    2024

  8. [16]

    D.; Re, C.; Acosta-Navas, D.; Hudson, D

    Liang, P.; Bommasani, R.; Lee, T.; Tsipras, D.; Soylu, D.; Yasunaga, M.; Zhang, Y.; Narayanan, D.; Wu, Y.; Kumar, A.; Newman, B.; Yuan, B.; Yan, B.; Zhang, C.; Cosgrove, C.; Manning, C. D.; Re, C.; Acosta-Navas, D.; Hudson, D. A.; Zelikman, E.; Durmus, E.; Ladhak, F.; Rong, F....

    2023

  9. [17]

    Lin, J.; Ma, X.; Lin, S.-C.; Yang, J.-H.; Pradeep, R.; and Nogueira, R. 2021. Pyserini: A Python Toolkit for Reproducible Information Retrieval Research with Sparse and Dense Representations. In Proceedings of the 44th International ACM SIGIR Conference on Research and Develop...

    2021

  10. [18]

    Liu, J.; Meng, S.; Gao, Y.; Mao, S.; Cai, P.; Yan, G.; Chen, Y.; Bian, Z.; Shi, B.; and Wang, D. 2025. Aligning vision to language: Text-free multimodal knowledge graph construction for enhanced llms reasoning. arXiv preprint arXiv:2503.12972

    2025

  11. [19]

    Ma, G.; Ma, Y.; Wu, X.; Su, Z.; Zhou, M.; and Hu, S. 2025. Task-level Distributionally Robust Optimization for Large Language Model-based Dense Retrieval. Proceedings of the AAAI Conference on Artificial Intelligence, 39(23): 24759--24767

    2025

  12. [20]

    Ma, X.; Wang, L.; Yang, N.; Wei, F.; and Lin, J. 2024. Fine-Tuning LLaMA for Multi-Stage Text Retrieval. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR '24, 2421–2425. Association for Computing Machiner...

    2024

  13. [21]

    Ma, X.; Zhang, X.; Pradeep, R.; and Lin, J. 2023. Zero-shot listwise document reranking with a large language model. arXiv preprint arXiv:2305.02156

    2023 arXiv

  14. [22]

    Meng, C.; Arabzadeh, N.; Askari, A.; Aliannejadi, M.; and de Rijke, M. 2024. Ranked List Truncation for Large Language Model-based Re-Ranking. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR '24, 141–151...

    2024

  15. [23]

    Nogueira, R.; Jiang, Z.; Pradeep, R.; and Lin, J. 2020 a . Document Ranking with a Pretrained Sequence-to-Sequence Model. In Cohn, T.; He, Y.; and Liu, Y., eds., Findings of the Association for Computational Linguistics: EMNLP 2020, 708--718. Online: Association for Computatio...

    2020

  16. [24]

    Nogueira, R.; Jiang, Z.; Pradeep, R.; and Lin, J. 2020 b . Document Ranking with a Pretrained Sequence-to-Sequence Model. In Cohn, T.; He, Y.; and Liu, Y., eds., Findings of the Association for Computational Linguistics: EMNLP 2020, 708--718. Online: Association for Computatio...

    2020

  17. [25]

    Podolak, J.; Peri\' c , L.; Jani\' c ijevi\' c , M.; and Petcu, R. 2025. Beyond Reproducibility: Advancing Zero-shot LLM Reranking Efficiency with Setwise Insertion. In Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrie...

    2025

  18. [26]

    Pradeep, R.; Sharifymoghaddam, S.; and Lin, J. 2023 a . RankVicuna: Zero-Shot Listwise Document Reranking with Open-Source Large Language Models. ArXiv, abs/2309.15088

    2023 arXiv

  19. [27]

    Pradeep, R.; Sharifymoghaddam, S.; and Lin, J. J. 2023 b . RankZephyr: Effective and Robust Zero-Shot Listwise Reranking is a Breeze! ArXiv, abs/2312.02724

    2023 arXiv

  20. [28]

    Qian, H.; Liu, Z.; Zhang, P.; Mao, K.; Lian, D.; Dou, Z.; and Huang, T. 2024. MemoRAG: Boosting Long Context Processing with Global Memory-Enhanced Retrieval Augmentation. Proceedings of the ACM on Web Conference 2025

    2024

  21. [29]

    Qin, Z.; Jagerman, R.; Hui, K.; Zhuang, H.; Wu, J.; Yan, L.; Shen, J.; Liu, T.; Liu, J.; Metzler, D.; Wang, X.; and Bendersky, M. 2024. Large Language Models are Effective Text Rankers with Pairwise Ranking Prompting. In Duh, K.; Gomez, H.; and Bethard, S., eds., Findings of t...

    2024

  22. [30]

    E.; Walker, S.; Jones, S.; Hancock-Beaulieu, M

    Robertson, S. E.; Walker, S.; Jones, S.; Hancock-Beaulieu, M. M.; Gatford, M.; et al. 1995. Okapi at TREC-3. British Library Research and Development Department

    1995

  23. [31]

    G.; Gaussier, E.; Li, M.; and Renders, J.-M

    Thonet, T.; Cinar, Y. G.; Gaussier, E.; Li, M.; and Renders, J.-M. 2022. Listwise Learning to Rank Based on Approximate Rank Indicators. Proceedings of the AAAI Conference on Artificial Intelligence, 36(8): 8494--8502

    2022

  24. [32]

    Trivedi, H.; Balasubramanian, N.; Khot, T.; and Sabharwal, A. 2022. MuSiQue: Multihop Questions via Single-hop Question Composition. Transactions of the Association for Computational Linguistics, 10: 539--554

    2022

  25. [33]

    Voorhees, E. M. 1999. Natural Language Processing and Information Retrieval. In Pazienza, M. T., ed., Information Extraction, 32--48. Berlin, Heidelberg: Springer Berlin Heidelberg. ISBN 978-3-540-48089-1

    1999

  26. [34]

    X.; Tu, X.; Wang, J.; Huang, A

    Wang, J.; Huang, J. X.; Tu, X.; Wang, J.; Huang, A. J.; Laskar, M. T. R.; and Bhuiyan, A. 2024. Utilizing BERT for Information Retrieval: Survey, Applications, Resources, and Challenges. ACM Comput. Surv., 56(7)

    2024

  27. [35]

    Xia, Y.; Zhou, J.; Shi, Z.; Chen, J.; and Huang, H. 2025. Improving Retrieval Augmented Language Model with Self-Reasoning. Proceedings of the AAAI Conference on Artificial Intelligence, 39(24): 25534--25542

    2025

  28. [36]

    W.; Salakhutdinov, R.; and Manning, C

    Yang, Z.; Qi, P.; Zhang, S.; Bengio, Y.; Cohen, W. W.; Salakhutdinov, R.; and Manning, C. D. 2018. HotpotQA: A dataset for diverse, explainable multi-hop question answering. arXiv preprint arXiv:1809.09600

    2018 arXiv

  29. [37]

    Yates, A.; Nogueira, R.; and Lin, J. 2021. Pretrained Transformers for Text Ranking: BERT and Beyond. In Proceedings of the 14th ACM International Conference on Web Search and Data Mining, WSDM '21, 1154–1156. New York, NY, USA: Association for Computing Machinery. ISBN 9781450382977

    2021

  30. [38]

    Yu, L.; Zhang, C.; Pei, S.; Sun, G.; and Zhang, X. 2018. WalkRanker: A Unified Pairwise Ranking Model With Multiple Relations for Item Recommendation. Proceedings of the AAAI Conference on Artificial Intelligence, 32(1)

    2018

  31. [39]

    Yu, Y.; Ping, W.; Liu, Z.; Wang, B.; You, J.; Zhang, C.; Shoeybi, M.; and Catanzaro, B. 2024. RankRAG: Unifying Context Ranking with Retrieval-Augmented Generation in LLMs. ArXiv, abs/2407.02485

    2024 arXiv

  32. [40]

    Q.; Guo, W.; Wang, H.; Li, Y.; Guo, H.; Liu, Y.; Lian, D.; and Chen, E

    Zhang, L.; Song, K.; Lee, Y. Q.; Guo, W.; Wang, H.; Li, Y.; Guo, H.; Liu, Y.; Lian, D.; and Chen, E. 2025 a . Killing Two Birds with One Stone: Unifying Retrieval and Ranking with a Single Generative Recommendation Model. In Proceedings of the 48th International ACM SIGIR Conf...

    2025

  33. [41]

    Zhang, L.; Zhang, Y.; Long, D.; Xie, P.; Zhang, M.; and Zhang, M. 2023. RankingGPT: Empowering Large Language Models in Text Ranking with Progressive Enhancement. CoRR

    2023

  34. [42]

    Zhang, Y.; Li, M.; Long, D.; Zhang, X.; Lin, H.; Yang, B.; Xie, P.; Yang, A.; Liu, D.; Lin, J.; Huang, F.; and Zhou, J. 2025 b . Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models. arXiv:2506.05176

    2025 arXiv

  35. [43]

    Zhuang, H.; Qin, Z.; Jagerman, R.; Hui, K.; Ma, J.; Lu, J.; Ni, J.; Wang, X.; and Bendersky, M. 2023. RankT5: Fine-Tuning T5 for Text Ranking with Ranking Losses. In Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval...

    2023

  36. [44]

    Zhuang, S.; Zhuang, H.; Koopman, B.; and Zuccon, G. 2024. A Setwise Approach for Effective and Highly Efficient Zero-shot Ranking with Large Language Models. In Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval, SIG...

    2024

  37. [45]

    Zuccon, G.; Zhuang, S.; and Ma, X. 2025. R2LLMs: Retrieval and Ranking with LLMs. In Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR '25, 4106–4109. New York, NY, USA: Association for Computing Machinery. I...

    2025

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.