Pith. sign in

REVIEW 3 major objections 6 minor 37 references

LESER: Learning to Expand via Search Engine-feedback Reinforcement in e-Commerce

T0 review · 3 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read LESER claims that fine-tuning a context-aware LLM with Group Relative Policy Optimization on live search-engine feedback yields query expansions that substantially improve retrieval coverage, relevance, and online engagement in e-commerce…

desk verdict A credible industrial GRPO-based query expansion method with a real circularity problem in offline evaluation; the A/B test is the only independent evidence and its numbers are confidential. read the letter →

arxiv 2509.05570 v2 pith:J2YSWK4I submitted 2025-09-06 cs.IR

Yipeng Zhang , Bowen Liu , Xiaoshuang Zhang , Aritra Mandal , Canran Xu , Zhe Wu This is my paper · ORCID
classification cs.IR
keywords queryexpansione-commercesearchreinforcementlearningGRPOlargelanguagemodelspseudo-relevancefeedbackretrievalcoverageonlineA/Btesting
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

LESER is a training framework for query expansion in e-commerce search. The paper claims that fine-tuning a context-aware LLM with Group Relative Policy Optimization, using real-time search-engine feedback as the reward, produces expansions that retrieve more relevant and more diverse products than prompt-based LLM expansion or supervised query rewriting. The method removes the need for labeled expansion data and grounds the model in the actual catalog through a pseudo-relevance-feedback context set. Offline results on 10,000 real queries show the 3B model improves retrieval coverage for 72.84% of queries and relevance for 32.40%, while a production A/B test reports statistically significant engagement gains. If the claim holds, this is an annotation-free route to query understanding that can scale to production traffic.

What carries the argument

The load-bearing machinery is the reward function of Eq. (1), $r(\mathcal{Y}) = r_{\mathrm{rel}}(\mathcal{Y}) + \lambda \cdot r_{\mathrm{size}}(\mathcal{Y})$ for valid outputs and $0$ otherwise, where $r_{\mathrm{rel}}$ normalizes the average relevance of the top-$k$ combined retrieval by the original query's average relevance and $r_{\mathrm{size}}$ normalizes the number of unique retrieved items by the original query's count. This reward turns the search engine into an oracle: the LLM samples $N$ expansion sets, each set is executed against the engine, and GRPO assigns relative advantages based on reward differences within the group. The validity gate enforces structural format, supported attributes, and length limits, while a supervised warm-up on distilled reasoning examples is required to keep the model emitting parseable `<think>`/`<answer>` JSON.

What would settle it

Re-run the offline evaluation with human relevance judgments in place of the GBDT model, scoring the newly retrieved items $R_{\mathrm{exp}} \setminus R_{\mathrm{orig}}$; if LESER's preferred expansions do not also raise human-judged relevance, or if the GBDT and human rankings diverge sharply on the expanded result set, the central claim would be refuted.

Watch

Extended reading notes

Core claim

The central claim is that query expansion for e-commerce search can be learned as a retrieval-optimization task: a context-aware LLM is fine-tuned with Group Relative Policy Optimization (GRPO) using a reward computed from live search-engine results, namely normalized relevance gain plus retrieval-coverage gain over the union of original and expanded queries. LESER grounds the LLM in a pseudo-relevance-feedback context set of top products with structured attributes, and a validity gate assigns zero reward to malformed output, keeping generations platform-compatible. On 10,000 real queries, the authors report that LESER with a 3B backbone yields positive retrieval gain for 72.84% of queries and positive relevance gain for 32.40%, exceeding embedding-based expansion, few-shot prompting of a proprietary LLM, and supervised fine-tuning baselines. A production A/B test on roughly 600K cached high-frequency queries showed statistically significant improvements in click-through rate, reduced query abandonment, and growth in gross merchandise value, with specific lift percentages kept confidential.

Load-bearing premise

The load-bearing premise is that the internal GBDT relevance model gives a faithful proxy for what users actually find relevant; if expansions can inflate that model's scores without improving real relevance, the offline gains and the RL objective are both misaligned with user satisfaction.

Editorial extensions

If this is right

  • Platforms with an existing relevance model can train expansion LLMs without human annotation, since the reward comes directly from live retrieval signals.
  • Because the reward is computed over the union of original plus expanded queries, the learned policy should favor expansions that add genuinely new relevant items rather than re-ranking the same result set.
  • The validity gate and warm-up stage suggest the approach transfers to other structured generation tasks where outputs must respect strict formats, such as facet extraction or filter inference.
  • With a 3B backbone fitting production latency budgets, similar RL-from-search-feedback training can be applied to higher-frequency or long-tail queries beyond the cached set.
  • The relative formulation, normalizing by the original query's own retrieval quality, makes the reward comparable across queries of very different specificity.

Reading between the lines

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

  • Editorial inference: the method's quality ceiling is set by the GBDT relevance model that provides the reward; if that model can be gamed by expansions that score well without reflecting true user relevance, the offline gains and the learned policy's objective would diverge from user satisfaction.
  • Editorial inference: the reward does not explicitly penalize overlap among the generated expansions, so the model could settle into near-duplicate queries that inflate the retrieval-size term without adding true diversity; a diversity-aware penalty would be a testable extension.
  • Editorial inference: the same search-engine-in-the-loop objective should transfer to other retrieval-side tasks such as attribute disambiguation, listing title rewriting, or query-to-filter prediction, wherever a catalog and a relevance model already exist.
  • Editorial inference: because the relevance model receives incremental updates over time, the reward distribution shifts, so LESER would likely need periodic re-training to stay aligned with the current relevance model.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper proposes LESER, a query-expansion framework that fine-tunes a context-aware LLM (LLaMA 3.2-3B) with Group Relative Policy Optimization (GRPO) using real-time search-engine feedback as the reward signal. The reward in Eq. (1) combines a normalized relevance metric and a retrieval-coverage metric computed from the search engine's top-k results. The authors report offline results on 10,000 real-world e-commerce queries, ablation studies over backbone sizes and training stages, qualitative examples, and a production A/B test. The central claim is that LESER substantially improves retrieval coverage and relevance relative to embedding-based expansion, few-shot GPT-4.1, and supervised fine-tuning, and delivers measurable user-engagement gains online. The paper positions LESER as a scalable replacement for prompt-based LLM expansion and supervised query rewriting in e-commerce search.

Significance. If the claims are validated, LESER is a practically meaningful contribution: it removes the need for supervised labels, incorporates live retrieval signals into LLM training, and demonstrates a production deployment path with a real A/B test. The use of search-engine feedback as a reward is a sensible formulation, and the GRPO-based relative-reward design is appropriate for the one-to-many nature of query expansion. The comparison against GPT-4.1 few-shot prompting and an embedding baseline is useful. However, the current evidence for the central claim is weakened by the fact that the offline reward and the offline evaluation metric share the same internal relevance model, and by the confidentiality of the online lift figures. The contribution is therefore defensible but not yet fully supported.

major comments (3)
  1. [§5.1, §5.2, Eq. (1)] The reward in Eq. (1) is computed from Rel(Y), which uses relevance scores from the internal GBDT relevance model described in Section 5.1, and the offline evaluation metric AvgRel in Section 5.2 is computed with the same model. The training objective therefore directly maximizes the quantity used to measure success, so the offline ΔRel gains are at least partially a fixed point of the optimization and may reflect gaming of the model rather than genuine relevance improvement. The paper reports no calibration of GBDT scores against human labels on expanded queries and no analysis of reward hacking. To support the relevance claim, the authors should provide an evaluation that does not share the reward model, such as human relevance judgments on the union retrieval set or engagement signals on the retrieved items, and report the correlation between GBDT scores and human labels for query expansions.
  2. [§5.4] The online A/B test is the only independent evidence for real-world efficacy, but Section 5.4 states that 'specific lift percentages remain confidential' and provides no confidence intervals, p-values, or details of the statistical protocol. A statement that results are 'statistically significant' is not verifiable by the reader. At a minimum, the paper should report significance levels, confidence intervals for relative changes, or an anonymized summary that does not reveal absolute revenue figures. Without such information, the central claim of 'measurable gains in user engagement' cannot be independently assessed.
  3. [§5.2, Table 2] The metrics ΔRet and ΔRel are defined in Section 5.2 as magnitudes (number of new unique items and difference in average relevance), but Table 2 reports the percentage of queries exhibiting any positive gain. This reporting choice conflates tiny, perhaps noisy, gains with substantial ones: a query gaining one marginal item is counted the same as a query gaining hundreds, and a relevance increase of 0.001 is counted the same as a large improvement. The authors should report the distribution of gains (mean, median, quantiles), the fraction of queries with gains above a meaningful threshold, and a query-level significance test to demonstrate that the improvements are not driven by small or noisy effects.
minor comments (6)
  1. [§5.2] In Section 5.2, ΔRet is defined as the cardinality of a set difference, but the reported value in Table 2 is a percentage of queries; the relationship between these two notions should be stated explicitly.
  2. [§1] There are typographical errors such as 'with in the expansion field' and 'an LLM is trained to perform context-aware reasoning'; these should be corrected in a revision.
  3. [§5.1] The sentence 'the relevance score is generated from internal relevance model from retrieved top 10 items' is grammatically unclear; it should specify which model produces the score and whether the top-10 context is the same set used for reward computation.
  4. [Table 3] In Table 3, the first row (no WarmUp, no GRPO) should be explicitly labeled as the base pretrained LLaMA 3.2-3B-Instruct model to avoid ambiguity about the setting.
  5. [Appendix A] Appendix A repeats the implementation details from Section 5.1 almost verbatim; the duplication should be removed or condensed to a summary.
  6. [References] The manuscript contains placeholder conference information ('Conference acronym ’XX, Woodstock, NY') and missing DOIs for several references; these should be completed before publication.

Circularity Check

1 steps flagged · score 5.0 of 10

Offline relevance and retrieval gains are measured by the same internal GBDT model that supplies the Eq. (1) reward, so the public offline evidence largely re-measures the training objective.

  1. fitted input called prediction [Section 5.1 (reward calculation) and Section 5.2 (Evaluation Metric)]
    "In the reward calculation, the relevance score is generated from internal relevance model from retrieved top 10 items. ... Let Rorig be its set of retrieved items and AvgRelorig be the average relevance score of its top 100 results. Similarly, let Rexp denote the union of items retrieved from both the original query and its corresponding expansions, with AvgRelexp as its associated average relevance score."

    The reward r_rel(Y) in Eq. (1) is defined from Rel(Y), the GBDT model's average relevance score of top-k items, and the offline metric ΔRel in Section 5.2 is the difference of the same GBDT model's average relevance (over top-100). LESER is trained by GRPO to maximize this reward, so the reported offline 'relevance gain' is a re-measurement of the optimization objective, not an independent signal of true relevance. No human-rated or externally calibrated relevance evaluation is reported for expanded queries, and online lift percentages are explicitly confidential (Section 5.4). Thus the offline relevance improvement is forced by the training objective to a substantial degree.

full rationale

The paper's central methodological contribution — using GRPO with live search-engine feedback (relevance and retrieval size) as reward — is not itself circular. The reward is computed from an internal GBDT relevance model (following [26]) and the offline evaluation metric ΔRel is computed from the same model, so the offline relevance gain largely re-measures the training objective rather than validating it against an independent ground truth. The retrieval gain ΔRet also mirrors the r_size reward component. The paper does not supply a human-rated or otherwise external check for expanded-query relevance, and the online A/B lift figures are confidential, leaving the public quantitative case substantially dependent on the non-independent offline metric. This is a partial circularity (reward and evaluation share the same fitted model), but the presence of the online experiment as independent evidence, though under-reported, prevents a score above 6. There are no load-bearing self-citations or imported uniqueness theorems.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The main unpaid-for inputs are the internal relevance model, the production search engine's behavior, and the design constants lambda, beta, epsilon, and context size. No new physical or synthetic entities are introduced.

free parameters (4)
  • lambda (reward trade-off) = 0.1
    Balances relevance gain against retrieval size gain in Eq. (1); chosen by hand, no sensitivity analysis reported.
  • beta (KL regularization weight) = 0.04
    Controls KL penalty in the GRPO objective, Eq. (3); chosen by hand.
  • epsilon (GRPO clipping range) = not stated
    Clipping range in the GRPO objective, Eq. (3), is not specified in the paper.
  • context size k = 10
    Context set and reward relevance scores use the top 10 retrieved items; a design choice without sensitivity analysis.
assumptions (4)
  • domain assumption The internal GBDT relevance model's continuous scores are a faithful proxy for human relevance judgments.
    Used as reward in Eq. (1) and as evaluation metric AvgRel in Section 5.2; if biased, offline gains may not reflect user-relevant quality.
  • domain assumption The production search engine's retrieval results used during training are representative of online traffic and stable across the evaluation period.
    Search-engine-in-the-loop training and offline evaluation both assume the engine's behavior at train time matches deployment.
  • domain assumption Top-10 retrieved items from the original query form a sufficient pseudo-relevant context for intent inference.
    Section 5.1 limits the context to no more than 10 items; no analysis of sensitivity to this choice is provided.
  • standard math Group Relative Policy Optimization (GRPO) converges to a policy that improves retrieval metrics.
    GRPO from [27] is adopted as a standard RL algorithm; no convergence guarantee is provided for this reward.

how reviews work

0 comments
Cite this review

Pith. "Pith review of LESER: Learning to Expand via Search Engine-feedback Reinforcement in e-Commerce." pith.science (2026). https://pith.science/paper/J2YSWK4I

@misc{pith2026250905570,
  author       = {Pith},
  title        = {Pith review of: LESER: Learning to Expand via Search Engine-feedback Reinforcement in e-Commerce},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J2YSWK4I}},
  note         = {Machine review of arXiv:2509.05570}
}
read the original abstract

User queries in e-commerce search are often vague, short, and underspecified, making it difficult for retrieval systems to match them accurately against structured product catalogs. This challenge is amplified by the one-to-many nature of user intent, where a single query can imply diverse and competing needs. Existing methods, including neural query expansion and prompting-based LLM approaches, fall short in real-world settings: they struggle to capture nuanced user intent, often generate outputs that violate platform constraints, and rely on workflows that are difficult to scale in production. We propose Learning to Expand via Search Engine-feedback Reinforcement (LESER), a novel framework that fine-tunes a context-aware LLM using real-time search engine feedback as supervision. LESER formulates query expansion as a retrieval optimization task and leverages Group Relative Policy Optimization to learn directly from relevance and coverage metrics. LESER is trained to reason over search results and produce high quality query expansions that align with platform rules and retrieval objectives. We evaluate LESER on large-scale, real-world e-commerce datasets, demonstrating substantial improvements in both offline and online settings. Our results show that LESER not only enhances semantic coverage and retrieval relevance but also delivers measurable gains in user engagement, making it a practical and scalable solution for modern search systems.

Figures

Figures reproduced from arXiv: 2509.05570 by the authors.

Figure 1
Figure 1. LESER Overview. Given a user query, a context set is retrieved from the search engine. The LLM then performs [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 15 canonical work pages

  1. [1]

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Floren- cia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. 2023. Gpt-4 technical report.arXiv preprint arXiv:2303.08774 (2023)

    arXiv 2023

  2. [2]

    Claudio Carpineto and Giovanni Romano. 2012. A survey of automatic query expansion in information retrieval.Acm Computing Surveys (CSUR)44, 1 (2012), 1–50

    2012

  3. [3]

    Lei Huang, Weijiang Yu, Weitao Ma, Weihong Zhong, Zhangyin Feng, Haotian Wang, Qianglong Chen, Weihua Peng, Xiaocheng Feng, Bing Qin, et al. 2025. A survey on hallucination in large language models: Principles, taxonomy, chal- lenges, and open questions.ACM Transactions on Information Systems43, 2 (2025), 1–55

    2025

  4. [4]

    Tuan-Luc Huynh, Thuy-Trang Vu, Weiqing Wang, Yinwei Wei, Trung Le, Dra- gan Gašević, Yuan-Fang Li, and Thanh-Toan Do. 2024. PromptDSI: Prompt- based Rehearsal-free Instance-wise Incremental Learning for Document Retrieval. arXiv:2406.12593 [cs.IR]

    arXiv 2024

  5. [5]

    Rolf Jagerman, Honglei Zhuang, Zhen Qin, Xuanhui Wang, and Michael Bender- sky. 2023. Query expansion by prompting large language models.arXiv preprint arXiv:2305.03653(2023)

    arXiv 2023

  6. [6]

    Anjishnu Kumar, Arpit Gupta, Julian Chan, Sam Tucker, Bjorn Hoffmeister, Markus Dreyer, Stanislav Peshterliev, Ankur Gandhe, Denis Filiminov, Ariya Rastrow, et al. 2017. Just ASK: building an architecture for extensible self-service spoken language understanding.arXiv preprint arXiv:1711.00549(2017). LESER: Learning to Expand via Search Engine-feedback Re...

    arXiv 2017

  7. [7]

    Bruce Croft

    Victor Lavrenko and W. Bruce Croft. 2001. Relevance-Based Language Models. InProceedings of the 24th Annual International ACM SIGIR Conference. 120–127

    work page 2001

  8. [8]

    Yibin Lei, Yu Cao, Tianyi Zhou, Tao Shen, and Andrew Yates. 2024. Corpus- Steered Query Expansion with Large Language Models. InProceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 2: Short Papers). 393–401

    work page 2024

Show all 37 references
  1. [9]

    Hang Li, Shengyao Zhuang, Bevan Koopman, and Guido Zuccon. 2025. LLM- VPRF: Large Language Model Based Vector Pseudo Relevance Feedback.arXiv preprint arXiv:2504.01448(2025)

    2025 arXiv

  2. [10]

    Aritra Mandal, Ishita K Khan, and Prathyusha Senthil Kumar. 2019. Query Rewriting using Automatic Synonym Extraction for E-commerce Search.. In eCOM@ SIGIR

    2019

  3. [11]

    Aritra Mandal, Daniel Tunkelang, and Zhe Wu. 2023. Semantic Equivalence of e-Commerce Queries.arXiv preprint arXiv:2308.03869(2023)

    2023 arXiv

  4. [12]

    Tran, Jinfeng Rao, Marc Najork, Emma Strubell, and Donald Metzler

    Sanket Vaibhav Mehta, Jai Gupta, Yi Tay, Mostafa Dehghani, Vinh Q. Tran, Jinfeng Rao, Marc Najork, Emma Strubell, and Donald Metzler. 2022. DSI++: Updating Transformer Memory with New Documents. arXiv:2212.09744 [cs.IR]

    2022 arXiv

  5. [13]

    Meta AI. 2024. Introducing Llama 3.1: Our most capable models to date. https: //ai.meta.com/blog/meta-llama-3-1/. Published July 23, 2024

    2024

  6. [14]

    Meta AI. 2024. Llama 3.2: Revolutionizing edge AI and vision with open, customiz- able models. https://ai.meta.com/blog/llama-3-2-connect-2024-vision-edge- mobile-devices/. Published September 25, 2024

    2024

  7. [15]

    Shahrzad Naseri, Jeffrey Dalton, Andrew Yates, and James Allan. 2021. Ceqe: Contextualized embeddings for query expansion. InAdvances in Information Retrieval: 43rd European Conference on IR Research, ECIR 2021, Virtual Event, March 28–April 1, 2021, Proceedings, Part I 43. Sp...

    2021

  8. [16]

    Petru Neague, Marcel Gregoriadis, and Johan Pouwelse. 2024. De-DSI: Decen- tralised Differentiable Search Index. arXiv:2404.12237 [cs.IR]

    2024 arXiv

  9. [17]

    Duy A Nguyen, Rishi Kesav Mohan, Van Yang, Pritom Saha Akash, and Kevin Chen-Chuan Chang. 2025. RL-based Query Rewriting with Distilled LLM for online E-Commerce Systems.arXiv preprint arXiv:2501.18056(2025)

    2025

  10. [18]

    Rodrigo Nogueira and Kyunghyun Cho. 2019. Document Expansion by Query Prediction. InProceedings of the 57th Annual Meeting of the Association for Compu- tational Linguistics (ACL) – Workshop on Question Answering for Complex Domains. 231–241

    2019

  11. [19]

    Rodrigo Nogueira, Weihang Yu, Hyun Joong Jo, and Kyunghyun Cho

  12. [20]

    Chester Palen-Michel, Ruixiang Wang, Yipeng Zhang, David Yu, Canran Xu, and Zhe Wu. 2024. Investigating LLM Applications in E-Commerce. arXiv:2408.12779 [cs.CL]

    2024 arXiv

  13. [21]

    Ming Pang, Chunyuan Yuan, Xiaoyu He, Zheng Fang, Donghao Xie, Fanyi Qu, Xue Jiang, Changping Peng, Zhangang Lin, Zheng Luo, and Jingping Shao. 2025. Generative Retrieval and Alignment Model: A New Paradigm for E-commerce Retrieval. arXiv:2504.01403 [cs.IR]

    2025 arXiv

  14. [22]

    Wenjun Peng, Guiyang Li, Yue Jiang, Zilong Wang, Dan Ou, Xiaoyi Zeng, Derong Xu, Tong Xu, and Enhong Chen. 2024. Large Language Model based Long-tail Query Rewriting in Taobao Search. InCompanion Proceedings of the ACM on Web Conference 2024, WWW 2024, Singapore, Singapore, Ma...

    2024

  15. [23]

    Samyam Rajbhandari, Jeff Rasley, Olatunji Ruwase, and Yuxiong He. 2020. Zero: Memory optimizations toward training trillion parameter models. InSC20: Inter- national Conference for High Performance Computing, Networking, Storage and Analysis. IEEE, 1–16

    2020

  16. [24]

    J. J. Rocchio. 1971. Relevance Feedback in Information Retrieval. InThe SMART Retrieval System—Experiments in Automatic Document Processing, Gerard Salton (Ed.). Prentice Hall, 313–323

    1971

  17. [25]

    Dwaipayan Roy, Debjyoti Paul, Mandar Mitra, and Utpal Garain. 2016. Using word embeddings for automatic query expansion.arXiv preprint arXiv:1606.07608 (2016)

    2016 arXiv

  18. [26]

    Dipanwita Saha, Anis Zaman, Hua Zou, Ning Chen, Xinxin Shu, Nadia Vase, and Abraham Bagherjeiran. 2025. Improving Ad matching via Cluster-Adaptive Key- word Expansion and Relevance tuning.arXiv preprint arXiv:2505.18897(2025)

    2025 arXiv

  19. [27]

    Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, YK Li, Y Wu, et al . 2024. Deepseekmath: Pushing the limits of mathematical reasoning in open language models.arXiv preprint arXiv:2402.03300(2024)

    2024 arXiv

  20. [28]

    Tran, Mostafa Dehghani, Jianmo Ni, Dara Bahri, Harsh Mehta, Zhen Qin, Kai Hui, Zhe Zhao, Jai Gupta, Tal Schuster, William W

    Yi Tay, Vinh Q. Tran, Mostafa Dehghani, Jianmo Ni, Dara Bahri, Harsh Mehta, Zhen Qin, Kai Hui, Zhe Zhao, Jai Gupta, Tal Schuster, William W. Cohen, and Donald Metzler. 2022. Transformer Memory as a Differentiable Search Index. arXiv:2202.06991 [cs.IR]

    2022 arXiv

  21. [29]

    Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, Aurelien Rodriguez, Armand Joulin, Edouard Grave, and Guillaume Lam- ple. 2023. LLaMA: Open and Efficient Foundation ...

    2023 arXiv

  22. [30]

    Liang Wang, Nan Yang, and Furu Wei. 2023. Query2doc: Query expansion with large language models.arXiv preprint arXiv:2303.07678(2023)

    2023 arXiv

  23. [31]

    Xiao Wang, Craig Macdonald, Nicola Tonellotto, and Iadh Ounis. 2021. Pseudo-Relevance Feedback for Multiple Representation Dense Retrieval. arXiv:2106.11251 [cs.IR]

    2021 arXiv

  24. [32]

    An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, Chujie Zheng, Dayiheng Liu, Fan Zhou, Fei Huang, Feng Hu, Hao Ge, Haoran Wei, Huan Lin, Jialong Tang, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jia...

    2025 arXiv

  25. [33]

    Wenjing Zhang, Zhaoxiang Liu, Kai Wang, and Shiguo Lian. 2024. Query Expan- sion and Verification with Large Language Model for Information Retrieval. In International Conference on Intelligent Computing. Springer, 341–351

    2024

  26. [34]

    Chenyu Zhao, Yunjiang Jiang, Yiming Qiu, Han Zhang, and Wen-Yun Yang. 2023. Differentiable Retrieval Augmentation via Generative Language Modeling for E-commerce Query Intent Classification. arXiv:2308.09308

    2023 arXiv

  27. [35]

    Gang Zhao, Ximing Zhang, Chenji Lu, Hui Zhao, Tianshu Wu, Pengjie Wang, Jian Xu, and Bo Zheng. 2025. Explainable LLM-driven Multi-dimensional Distillation for E-Commerce Relevance Learning. InCompanion Proceedings of the ACM on Web Conference 2025. 631–640

    2025

  28. [36]

    Yuze Zhao, Jintao Huang, Jinghan Hu, Xingjun Wang, Yunlin Mao, Daoze Zhang, Zeyinzi Jiang, Zhikai Wu, Baole Ai, Ang Wang, Wenmeng Zhou, and Yingda Chen. 2024. SWIFT:A Scalable lightWeight Infrastructure for Fine-Tuning. arXiv:2408.05517 [cs.CL] https://arxiv.org/abs/2408.05517...

    2024 arXiv

  29. [2020]

    arXiv:2004.14220 [cs.IR]

    Document Expansion and Answer Passage Retrieval Using GPT-2. arXiv:2004.14220 [cs.IR]

    2004 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.