Pith. sign in

REVIEW 4 major objections 5 minor 46 references

Towards a Unified Paradigm: Integrating Recommendation Systems as a New Language in Large Models

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The RSLLM framework, which merges item ID embeddings with text inside LLM prompts, reports top HitRatio@1 on MovieLens, Steam, and LastFM.

desk verdict A useful incremental recipe for fusing ID embeddings and text in LLM recommenders, but the claimed edge over LLaRA rests on borrowed baseline numbers and unverifiable statistics. read the letter →

arxiv 2412.16933 v1 pith:3W7WKNSS submitted 2024-12-22 cs.IR cs.AIcs.CL

Kai Zheng , Qingfeng Sun , Can Xu , Peng Yu , Qingwei Guo This is my paper · ORCID
classification cs.IRcs.AIcs.CL
keywords sequentialrecommendationlargelanguagemodelsasanewcontrastivealignmentitemIDembeddingshybridpromptingLoRAfine-tuningHitRatio@1
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 a large language model can become a better sequential recommender if it is taught a new 'language' that mixes textual item descriptions with the learned ID embeddings produced by a traditional recommender. The proposed framework, RSLLM, builds prompts in which each item appears both as its title and as a behavioral token, aligns the ID embeddings to the LLM's input space with a small projector, and fine-tunes the LLM in two stages using two contrastive losses plus a next-item prediction loss. The authors report that RSLLM outperforms all compared baselines on MovieLens, Steam, and LastFM, with HitRatio@1 gains of 4.3%, 3.0%, and 4.8% over the best baseline on each dataset. If the result is correct, it means the gap between traditional collaborative recommenders and LLM-based recommenders can be narrowed by giving the LLM behavioral signals in addition to text.

What carries the argument

The central mechanism is the hybrid token representation, which concatenates three components for every item: the LLM's own text tokens for the item title, a special '[DSR]' token marking a recommendation ID, and a behavioral token computed by an Adapter built from a traditional sequential recommender (GRU4Rec, Caser, or SASRec). A trainable projector, a two-layer perceptron, maps the recommender's ID embedding into the LLM's input space so the model can read the two kinds of information together. The training objective combines the next-item prediction loss with two InfoNCE-style contrastive losses, one aligning the user-history tower with the target-item tower and one aligning the target item conditioned on history with the target item alone. Two-stage fine-tuning first runs text-only prompts, then unified prompts, with LoRA used for parameter-efficient adaptation. This is what carries the argument: it lets the model use world knowledge from text and collaborative knowledge from ID embeddings at multiple granularities.

What would settle it

Re-run the MovieLens, Steam, and LastFM comparisons with the same candidate protocol but report per-run HitRatio@1 and standard deviations, and additionally evaluate under full-corpus ranking where the true next item must be ranked among all items instead of a sampled 20-item set. If RSLLM's margins over LLaRA shrink to within noise or invert under full-corpus ranking, the central claim would be weakened.

Watch

Extended reading notes

Core claim

The paper's central claim is that its RSLLM framework significantly outperforms all baseline models across the three datasets in HitRatio@1 and ValidRatio. On MovieLens, RSLLM surpasses the best baseline by 4.3% in HitRatio@1; on Steam the gain is 3.0%; on LastFM it is 4.8%. The authors attribute the improvement to the unified prompting method that combines ID-based item embeddings from a conventional sequential recommender with textual item features, and to a two-stage fine-tuning procedure that first adapts the LLM with text-only prompts and then with unified prompts, using two contrastive losses and a language modeling loss. Ablation experiments support the contribution of each component: removing textual features, item ID representation, IID tokens, or contrastive alignment each reduces performance.

Load-bearing premise

The reported superiority of RSLLM rests on the evaluation protocol from LLaRA: for each user sequence, 20 non-interacted items are randomly sampled as candidates, the model must pick the true next item from that set, and success is averaged over five runs without reporting standard deviations; if this candidate-based protocol is a biased measure of real next-item recommendation quality, the few-point gains over LLaRA may not transfer to production settings.

Editorial extensions

If this is right

  • LLM-based recommenders can incorporate collaborative signals by treating item IDs as a new language rather than as plain numbers or text alone.
  • Two-stage training matters: the text-only stage is a necessary foundation for the behavioral stage, and removing either stage reduces HitRatio@1.
  • Item-item contrastive alignment contributes more than user-item alignment, showing that the model benefits from aligning the predicted target with the actual target at the item level.
  • The framework works with RNN-, CNN-, and self-attention-based backbones, with SASRec giving the best results, so the gains come from the prompting and training scheme rather than a particular recommender.
  • Instruction tuning on the recommendation task lifts ValidRatio from 0.17-0.44 for the base Llama2 to above 0.96 for RSLLM, addressing the invalid-answer problem of generative recommenders.

Reading between the lines

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

  • A natural stress test is to move from the 20-candidate protocol to full-corpus ranking; if RSLLM's HitRatio@1 margins persist when the true item must be ranked among all items, the collaborative-signal hypothesis would be much stronger. (Editorial inference, not reported in the paper.)
  • The 'new language' idea is not specific to item IDs: the same projector-plus-two-stage-contrastive recipe could be applied to timestamps, user attributes, or multimodal signals, and the paper's framework suggests a general recipe for injecting structured knowledge into LLMs. (Editorial inference.)
  • Because no standard deviations are reported, an independent re-run with the same settings and multiple seeds would clarify whether the few-percent gains are consistent or within noise. (Editorial inference.)
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 RSLLM, a framework for sequential recommendation that integrates traditional ID-based recommender embeddings with LLM textual prompts. Items are represented in a hybrid prompt containing both textual metadata and behavioral tokens derived from collaborative filtering embeddings via a trainable projector (Section 4.1). Training is a two-stage LoRA fine-tuning procedure that combines a next-item language modeling loss with two contrastive losses (Eqs. 2-5). The method is evaluated on MovieLens, Steam, and LastFM, using HitRatio@1 and ValidRatio under a 20-item candidate-set protocol following LLaRA (Section 5.1). The paper claims significant improvement over all baselines, including traditional recommenders and LLM-based methods, with ablations supporting the contribution of textual features, ID information, contrastive alignment, and two-stage training.

Significance. If the empirical claims hold, RSLLM is a useful contribution to LLM-based sequential recommendation: it explicitly aligns traditional recommender embeddings with LLM token space at multiple granularities, and the two-stage contrastive fine-tuning scheme is clearly described and reproducible in principle. The consistent gains over strong baselines across three datasets and three recommender backbones are a positive signal. However, the paper's central claim is empirical and depends entirely on the controlled comparison with LLaRA and other baselines. The manuscript does not ship code or release per-run results, and the baseline numbers for LLaRA are borrowed rather than re-run in the same evaluation harness. These issues currently prevent verifying the claimed significance, although they are fixable within the scope of a revision.

major comments (4)
  1. [Section 5.2, Tables 2 and 4] The central claim that RSLLM 'significantly outperforms all baseline models' is not established by the reported evidence. In Table 4, the baseline row (♣) is explicitly labeled 'results taken from (Liao et al., 2024)', meaning the strongest LLM baseline, LLaRA, was not evaluated in the authors' own harness. The margins over LLaRA can be very small: on Steam, RSLLM-Caser achieves HitRatio@1 of 0.4953 versus LLaRA-SASRec's 0.4949 (Table 2), a difference of 0.0004. Given that HitRatio@1 is computed over randomly sampled 20-item candidate sets (Section 5.1), any difference in candidate sampling, prompt template, decoding settings, or sequence filtering between the two papers could shift the metric by more than this gap. The authors must re-run LLaRA and the other baselines under the identical candidate-set protocol and report the resulting numbers, or clearly demonstrate that borrowed numbers are directly comparable.
  2. [Section 5.1 and Table 2 caption] The paired t-test claim in the Table 2 caption is unverifiable. The paper reports only averages over five runs, with no standard deviations, no test statistics, no degrees of freedom, and no statement of whether the pairing is across runs or across sequences. A claim of 'significant best and second-best results' with p<0.05 requires at minimum per-run values or error bars and a clear description of the paired test setup. Without this, the reader cannot assess whether the observed improvements over LLaRA are statistically meaningful, especially for margins as small as those on Steam.
  3. [Section 5.3 and Table 4] The ablation study has the same controlled-comparison problem as the main results: the baseline row is borrowed from Liao et al. (2024), while the RSLLM rows are produced in this paper's harness. Moreover, the ablation text draws conclusions such as 'without pre-loading item embeddings, the model achieves similar performance as when the embeddings are included' without reporting standard deviations or significance tests. Given that the differences between adjacent rows are often small (e.g., 0.4890 vs. 0.4905 on Steam for w/o IID Tokens versus w/o Pre-loading), the ablation conclusions require the same statistical support as the main results. Please report error bars or per-run results for all ablation configurations.
  4. [Section 5.1, Baseline list] The baseline enumeration in Section 5.1 lists 'MetaST' as a baseline, but MetaST does not appear in Table 2 or anywhere in the results or appendices. Either provide its results under the same protocol or remove it from the baseline list. In addition, the setup does not state which hyperparameters or prompt templates were used for MoRec and TALLRec, nor whether these baselines were re-run by the authors or taken from prior work. The reader needs this information to judge the fairness of the comparison.
minor comments (5)
  1. [Section 4.2, Eq. (2)] The sentence preceding Eq. (2) says 'the first m tokens correspond to all items except the last one, with the remaining n−m tokens dedicated to the target item,' but the notation is ambiguous because the sequence is written as (u1, u2, ..., un) and m is not defined in terms of n or the number of items. Please clarify the token-level indexing.
  2. [Section 5.3, text near Table 4] The phrase 'removing item ID representation or IID tokens also delegate the performance' should read 'degrade the performance'. Also, the sentence 'It is recommended to directly train the item representation parameters' is confusing after the row 'w/o Pre-loading item Embeddings' shows similar performance; please clarify whether the recommendation is to train from scratch without preloading.
  3. [Section 5.4 and Figure 3 caption] The caption of Figure 3 lists 'LLaRA' among the compared representations, but the text describes only numerical indexing, behavior tokens, text feature, and RSLLM representation. Please make the figure caption consistent with the text and explain how LLaRA's representation is defined in this comparison.
  4. [Section 5.1, Implementation Details] The hyperparameters γ=0.3, β=0.4, and τc=0.5 are said to be determined by grid search, but no grid ranges or selection criterion are given. State the search ranges and whether the same tuning budget was applied to baselines, and report the selected hyperparameters per dataset if they vary.
  5. [General] The paper does not mention code or data release. Given that the empirical protocol (candidate sampling, prompt templates, sequence filtering) is central to the claims, releasing code and processed data, or at least detailed pseudocode for the evaluation, would substantially improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: RSLLM's reported gains are empirical test-set measurements, not consequences forced by the model construction or by a self-citation chain.

full rationale

RSLLM is an empirical system paper rather than a derivation. The central claim in Section 5.2 is that RSLLM outperforms baselines on HitRatio@1 and ValidRatio; those numbers are measured on held-out candidate sets under the protocol described in Section 5.1. The training objective in Eq. 5 combines a language-modeling loss with two contrastive losses, and the hyperparameters gamma, beta, and tau are chosen by grid search; optimizing these losses fits the training data but does not by construction determine the reported test-set HitRatio@1 values. The only potentially suspicious benchmarking ingredient is Table 4's baseline row, marked with a club symbol and described as 'results taken from (Liao et al., 2024)', meaning LLaRA numbers were not reproduced inside this paper's harness. That is a comparability and reproducibility caveat, not circularity: the borrowed baseline numbers are external inputs to the comparison, not outputs of this paper being fed back as assumptions. No load-bearing step relies on a uniqueness theorem or on a self-citation chain; the authors' earlier works appear only as background references. The phrase 'treats sequential behaviors as a distinct language' is a framing metaphor, not an equation that equates the conclusion with the premise. Accordingly, there is no step where the claimed result is assumed in its own construction.

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

The central claim rests on domain assumptions about contrastive transfer and prompt-based fine-tuning; there is no formal derivation. The hand-chosen scalars are the two loss weights, the contrastive temperature, and the candidate set size, with the latter two adopted or selected by grid search.

free parameters (4)
  • contrastive loss weight gamma = 0.3
    Set by grid search; controls item-item contrastive alignment relative to next-token loss.
  • contrastive loss weight beta = 0.4
    Set by grid search; controls user-item contrastive alignment.
  • InfoNCE temperature tau_c = 0.5
    Temperature in Eq. 3 and Eq. 4; chosen by grid search.
  • Candidate set size K = 20
    Evaluation setting from LLaRA; every reported HitRatio@1 is defined against 20 random non-interacted candidates, so the headline numbers depend on this choice.
assumptions (4)
  • domain assumption Pretrained LLM weights can be adapted via LoRA to consume projected ID embeddings without destroying language modeling ability.
    Section 4.2 assumes the hybrid prompts remain usable after LoRA tuning; no formal guarantee is given.
  • domain assumption InfoNCE-style contrastive losses improve item and user alignment for sequential recommendation.
    Equations 3-5 assume cosine-similarity contrastive learning transfers to this setting; the ablation supports but does not prove it.
  • domain assumption ID embeddings from GRU4Rec, Caser, and SASRec remain informative after projection into the LLM input space.
    Section 4.1.2 relies on this transfer; the paper does not analyze what information survives projection.
  • domain assumption The candidate set protocol of 20 random non-interacted items per sequence measures next-item prediction without bias.
    Section 5.1 adopts this protocol from LLaRA; the entire comparison is defined relative to it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards a Unified Paradigm: Integrating Recommendation Systems as a New Language in Large Models." pith.science (2026). https://pith.science/paper/3W7WKNSS

@misc{pith2026241216933,
  author       = {Pith},
  title        = {Pith review of: Towards a Unified Paradigm: Integrating Recommendation Systems as a New Language in Large Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3W7WKNSS}},
  note         = {Machine review of arXiv:2412.16933}
}
read the original abstract

This paper explores the use of Large Language Models (LLMs) for sequential recommendation, which predicts users' future interactions based on their past behavior. We introduce a new concept, "Integrating Recommendation Systems as a New Language in Large Models" (RSLLM), which combines the strengths of traditional recommenders and LLMs. RSLLM uses a unique prompting method that combines ID-based item embeddings from conventional recommendation models with textual item features. It treats users' sequential behaviors as a distinct language and aligns the ID embeddings with the LLM's input space using a projector. We also propose a two-stage LLM fine-tuning framework that refines a pretrained LLM using a combination of two contrastive losses and a language modeling loss. The LLM is first fine-tuned using text-only prompts, followed by target domain fine-tuning with unified prompts. This trains the model to incorporate behavioral knowledge from the traditional sequential recommender into the LLM. Our empirical results validate the effectiveness of our proposed framework.

Figures

Figures reproduced from arXiv: 2412.16933 by the authors.

Figure 1
Figure 1. An illustration of prior item representation methods and ours. (a) ID Number: represents an item with a [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. The performance comparison of different item [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 8 canonical work pages

  1. [1]

    Keqin Bao, Jizhi Zhang, Yang Zhang, Wenjie Wang, Fuli Feng, and Xiangnan He. 2023. https://doi.org/10.1145/3604915.3608857 Tallrec: An effective and efficient tuning framework to align large language model with recommendation . In Proceedings of the 17th ACM Conference on Recommender Systems, RecSys ’23. ACM

    arXiv 2023

  2. [2]

    Tom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert - Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel M. Ziegler, Jeffrey Wu, Clemens Winter, Christopher Hesse, Mark Chen, Eric Sigler, Mateusz Litw...

    arXiv 2020

  3. [3]

    Ivan Cantador, Peter Brusilovsky, and Tsvi Kuflik. 2011. https://doi.org/10.1145/2043932.2044016 Second workshop on information heterogeneity and fusion in recommender systems (hetrec2011) . In Proceedings of the Fifth ACM Conference on Recommender Systems, RecSys '11, page 387–388, New York, NY, USA. Association for Computing Machinery

    work page arXiv 2011

  4. [4]

    Ting Chen, Simon Kornblith, Mohammad Norouzi, and Geoffrey Hinton. 2020. https://proceedings.mlr.press/v119/chen20j.html A simple framework for contrastive learning of visual representations . In Proceedings of the 37th International Conference on Machine Learning, volume 119 of Proceedings of Machine Learning Research, pages 1597--1607. PMLR

    2020

  5. [5]

    Zeyu Cui, Jianxin Ma, Chang Zhou, Jingren Zhou, and Hongxia Yang. 2022. https://arxiv.org/abs/2205.08084 M6-rec: Generative pretrained language models are open-ended recommender systems . Preprint, arXiv:2205.08084

    arXiv 2022

  6. [6]

    Sunhao Dai, Ninglu Shao, Haiyuan Zhao, Weijie Yu, Zihua Si, Chen Xu, Zhongxiang Sun, Xiao Zhang, and Jun Xu. 2023. https://doi.org/10.1145/3604915.3610646 Uncovering chatgpt’s capabilities in recommender systems . In Proceedings of the 17th ACM Conference on Recommender Systems, RecSys ’23. ACM

    arXiv 2023

  7. [7]

    Hui Fang, Danning Zhang, Yiheng Shu, and Guibing Guo. 2019. https://arxiv.org/abs/1905.01997 Deep learning-based sequential recommender systems: Concepts, algorithms, and evaluations . CoRR, abs/1905.01997

    arXiv 2019

  8. [8]

    Tianyu Gao, Xingcheng Yao, and Danqi Chen. 2021. https://doi.org/10.18653/v1/2021.emnlp-main.552 S im CSE : Simple contrastive learning of sentence embeddings . In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 6894--6910, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics

    doi:10.18653/v1/2021.emnlp-main.552 2021

Show all 46 references
  1. [9]

    Shijie Geng, Shuchang Liu, Zuohui Fu, Yingqiang Ge, and Yongfeng Zhang. 2023. https://arxiv.org/abs/2203.13366 Recommendation as language processing (rlp): A unified pretrain, personalized prompt & predict paradigm (p5) . Preprint, arXiv:2203.13366

    2023 arXiv

  2. [10]

    Zhen Guo, Peiqi Wang, Yanwei Wang, and Shangdi Yu. 2023. Dr. llama: Improving small language models in domain-specific qa via generative data augmentation

    2023

  3. [11]

    Maxwell Harper, Joseph A

    F. Maxwell Harper, Joseph A. Konstan, and Joseph A. 2016. https://api.semanticscholar.org/CorpusID:16619709 The movielens datasets: History and context . ACM Trans. Interact. Intell. Syst., 5:19:1--19:19

    2016

  4. [12]

    Xingwei He, Zheng-Wen Lin, Yeyun Gong, Alex Jin, Hang Zhang, Chen Lin, Jian Jiao, Siu Ming Yiu, Nan Duan, and Weizhu Chen. 2023. Annollm: Making large language models to be better crowdsourced annotators. ArXiv, abs/2303.16854

    2023 arXiv

  5. [14]

    Bal \' a zs Hidasi, Alexandros Karatzoglou, Linas Baltrunas, and Domonkos Tikk. 2016. http://arxiv.org/abs/1511.06939 Session-based recommendations with recurrent neural networks . In 4th International Conference on Learning Representations, ICLR 2016, San Juan, Puerto Rico, M...

    2016 arXiv

  6. [15]

    Yupeng Hou, Zhankui He, Julian McAuley, and Wayne Xin Zhao. 2023. https://arxiv.org/abs/2210.12316 Learning vector-quantized item representation for transferable sequential recommenders . Preprint, arXiv:2210.12316

    2023 arXiv

  7. [16]

    Yupeng Hou, Junjie Zhang, Zihan Lin, Hongyu Lu, Ruobing Xie, Julian McAuley, and Wayne Xin Zhao. 2024. https://arxiv.org/abs/2305.08845 Large language models are zero-shot rankers for recommender systems . Preprint, arXiv:2305.08845

    2024 arXiv

  8. [17]

    Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen

    Edward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2021. https://arxiv.org/abs/2106.09685 Lora: Low-rank adaptation of large language models . Preprint, arXiv:2106.09685

    2021 arXiv

  9. [18]

    Zhiqiang Hu, Yihuai Lan, Lei Wang, Wanyu Xu, Ee-Peng Lim, Roy Ka-Wei Lee, Lidong Bing, and Soujanya Poria. 2023. Llm-adapters: An adapter family for parameter-efficient fine-tuning of large language models. ArXiv, abs/2304.01933

    2023 arXiv

  10. [19]

    Wenyue Hua, Shuyuan Xu, Yingqiang Ge, and Yongfeng Zhang. 2023. https://doi.org/10.1145/3624918.3625339 How to index item ids for recommendation foundation models . In Proceedings of the Annual International ACM SIGIR Conference on Research and Development in Information Retri...

    2023

  11. [20]

    Wang - Cheng Kang and Julian J. McAuley. 2018. https://arxiv.org/abs/1808.09781 Self-attentive sequential recommendation . CoRR, abs/1808.09781

    2018 arXiv

  12. [21]

    Jia Li, Ge Li, Yongming Li, and Zhi Jin. 2023 a . Enabling programming thinking in large language models toward code generation

    2023

  13. [22]

    Jiacheng Li, Ming Wang, Jin Li, Jinmiao Fu, Xin Shen, Jingbo Shang, and Julian McAuley. 2023 b . https://arxiv.org/abs/2305.13731 Text is all you need: Learning language representations for sequential recommendation . Preprint, arXiv:2305.13731

    2023 arXiv

  14. [23]

    Yaoyiran Li, Ching-Yun Chang, Stephen Rawls, Ivan Vuli \'c , and Anna Korhonen. 2023 c . https://doi.org/10.18653/v1/2023.acl-long.510 Translation-enhanced multilingual text-to-image generation . In Proceedings of the 61st Annual Meeting of the Association for Computational Li...

    2023 doi

  15. [24]

    Yaoyiran Li, Fangyu Liu, Nigel Collier, Anna Korhonen, and Ivan Vuli \'c . 2022. https://doi.org/10.18653/v1/2022.acl-long.299 Improving word translation via two-stage contrastive learning . In Proceedings of the 60th Annual Meeting of the Association for Computational Linguis...

    2022 doi

  16. [25]

    Jiayi Liao, Sihang Li, Zhengyi Yang, Jiancan Wu, Yancheng Yuan, Xiang Wang, and Xiangnan He. 2024. https://arxiv.org/abs/2312.02445 Llara: Large language-recommendation assistant . Preprint, arXiv:2312.02445

    2024 arXiv

  17. [26]

    Junling Liu, Chao Liu, Peilin Zhou, Renjie Lv, Kang Zhou, and Yan Zhang. 2023. https://arxiv.org/abs/2304.10149 Is chatgpt a good recommender? a preliminary study . Preprint, arXiv:2304.10149

    2023 arXiv

  18. [27]

    Ziyang Luo, Can Xu, Pu Zhao, Xiubo Geng, Chongyang Tao, Jing Ma, Qingwei Lin, and Daxin Jiang. 2023. Augmented large language models with parametric knowledge guiding. ArXiv, abs/2305.04757

    2023 arXiv

  19. [28]

    Fu, Qinghua Hu, and Bing Wu

    Huan Ma, Changqing Zhang, Yatao Bian, Lemao Liu, Zhirui Zhang, Peilin Zhao, Shu Zhang, H. Fu, Qinghua Hu, and Bing Wu. 2023. Fairness-guided few-shot prompting for large language models. ArXiv, abs/2303.13217

    2023 arXiv

  20. [29]

    OpenAI, Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, Red Avila, Igor Babuschkin, Suchir Balaji, Valerie Balcom, Paul Baltescu, Haiming Bao, Mohammad Bavarian, Jeff ...

    2024 arXiv

  21. [30]

    Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. 2022. Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 3...

    2022

  22. [31]

    Kunzhe Song, Qingfeng Sun, Can Xu, Kai Zheng, and Yaming Yang. 2023. https://api.semanticscholar.org/CorpusID:258332030 Self-supervised multi-modal sequential recommendation . ArXiv, abs/2304.13277

    2023 arXiv

  23. [32]

    Jiaxi Tang and Ke Wang. 2018. https://arxiv.org/abs/1809.07426 Personalized top-n sequential recommendation via convolutional sequence embedding . CoRR, abs/1809.07426

    2018 arXiv

  24. [33]

    Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, W...

    2023 arXiv

  25. [34]

    A \" a ron van den Oord, Yazhe Li, and Oriol Vinyals. 2018. https://arxiv.org/abs/1807.03748 Representation learning with contrastive predictive coding . CoRR, abs/1807.03748

    2018 arXiv

  26. [35]

    Sheng, and Mehmet A

    Shoujin Wang, Liang Hu, Yan Wang, Longbing Cao, Quan Z. Sheng, and Mehmet A. Orgun. 2020. https://arxiv.org/abs/2001.04830 Sequential recommender systems: Challenges, progress and prospects . CoRR, abs/2001.04830

    2020 arXiv

  27. [36]

    Le Wu, Xiangnan He, Xiang Wang, Kun Zhang, and Meng Wang. 2022. A survey on accuracy-oriented neural recommendation: From collaborative filtering to information-rich recommendation. IEEE Transactions on Knowledge and Data Engineering, 35(5):4425--4445

    2022

  28. [37]

    Shijie Wu, Ozan Irsoy, Steven Lu, Vadim Dabravolski, Mark Dredze, Sebastian Gehrmann, Prabhanjan Kambadur, David Rosenberg, and Gideon Mann. 2023. https://arxiv.org/abs/2303.17564 Bloomberggpt: A large language model for finance . Preprint, arXiv:2303.17564

    2023 arXiv

  29. [38]

    Can Xu, Qingfeng Sun, Kai Zheng, Xiubo Geng, Pu Zhao, Jiazhan Feng, Chongyang Tao, and Daxin Jiang. 2023. https://api.semanticscholar.org/CorpusID:258298159 Wizardlm: Empowering large language models to follow complex instructions . ArXiv, abs/2304.12244

    2023 arXiv

  30. [39]

    Zheng Yuan, Fajie Yuan, Yu Song, Youhua Li, Junchen Fu, Fei Yang, Yunzhu Pan, and Yongxin Ni. 2023. https://arxiv.org/abs/2303.13835 Where to go next for recommender systems? id- vs. modality-based recommender models revisited . Preprint, arXiv:2303.13835

    2023 arXiv

  31. [40]

    An Zhang, Yuxin Chen, Leheng Sheng, Xiang Wang, and Tat-Seng Chua. 2024. https://arxiv.org/abs/2310.10108 On generative agents in recommendation . Preprint, arXiv:2310.10108

    2024 arXiv

  32. [41]

    Chen, Jinhao Jiang, Ruiyang Ren, Yifan Li, Xinyu Tang, Zikang Liu, Peiyu Liu, Jianyun Nie, and Ji rong Wen

    Wayne Xin Zhao, Kun Zhou, Junyi Li, Tianyi Tang, Xiaolei Wang, Yupeng Hou, Yingqian Min, Beichen Zhang, Junjie Zhang, Zican Dong, Yifan Du, Chen Yang, Yushuo Chen, Z. Chen, Jinhao Jiang, Ruiyang Ren, Yifan Li, Xinyu Tang, Zikang Liu, Peiyu Liu, Jianyun Nie, and Ji rong Wen. 20...

    2023 arXiv

  33. [42]

    Yuyue Zhao, Jiancan Wu, Xiang Wang, Wei Tang, Dingxian Wang, and Maarten de Rijke. 2024. https://api.semanticscholar.org/CorpusID:270045066 Let me do it for you: Towards llm empowered recommendation via tool learning

    2024

  34. [43]

    Kai Zheng, Qingfeng Sun, Yaming Yang, Tengchao Lv, Yeyong Pi, Changlin Zhao, Fei Xu, and Qi Zhang. 2023 a . https://api.semanticscholar.org/CorpusID:259859117 Adversarial knowledge stimulated contrastive prompting for few-shot language learners . In Annual Meeting of the Assoc...

    2023

  35. [44]

    Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, Dacheng Li, Eric Xing, et al. 2023 b . Judging llm-as-a-judge with mt-bench and chatbot arena. arXiv preprint arXiv:2306.05685

    2023 arXiv

  36. [45]

    Deyao Zhu, Jun Chen, Xiaoqian Shen, Xiang Li, and Mohamed Elhoseiny. 2023. Minigpt-4: Enhancing vision-language understanding with advanced large language models. ArXiv, abs/2304.10592

    2023 arXiv

  37. [46]

    online" 'onlinestring :=

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

  38. [47]

    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 11, 2026 · model on record in the stance chip above.