pith. sign in

arxiv: 2606.07995 · v1 · pith:W5WWONE4new · submitted 2026-06-06 · 💻 cs.CL

Customer-Agent: Overcoming Context Limitations in Ultra-Long Shopping Trajectories via Tool-Augmented Agents and RLVR

Hongye Liu , Rongmei Lin , Anurag Kashyap , Hejie Cui , Ricardo Henao , Besnik Fetahu , Bing Yin This is my paper

Pith reviewed 2026-06-27 20:03 UTC · model grok-4.3

classification 💻 cs.CL
keywords ShopTrajQAlong shopping trajectoriestool-augmented agentsRLVRcode interpreterLLM context limitscustomer behavior analysisagentic training
0
0 comments X

The pith

An RLVR-trained agent stores ultra-long shopping trajectories externally and retrieves them via code tools to bypass LLM context windows.

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

Real customer shopping records often cover multiple years and exceed the fixed context lengths of current large language models, limiting personalized analysis. The paper introduces the ShopTrajQA benchmark built from real product data and simulated trajectories with variants reaching 32k and 64k tokens to measure this gap. It presents a Customer Agent Framework that keeps trajectories in external files and trains the agent with Reinforcement Learning with Verifiable Rewards to generate and execute code-interpreter queries such as SQL. The central claim is that this external-tool approach enables accurate reasoning over arbitrarily long trajectories where direct context loading fails. If the method works, it supports handling genuine e-commerce histories without truncation or lossy summarization.

Core claim

The Customer Agent Framework stores trajectories as external local files and trains the agent via an RLVR paradigm to autonomously retrieve and parse them through code-interpreter interactions such as SQL queries, thereby bypassing the fixed in-context window constraints of LLMs while delivering strong performance on ShopTrajQA and generalizing to other complex reasoning tasks.

What carries the argument

The RLVR-trained Customer Agent that issues code-interpreter calls to query and analyze externally stored trajectory files.

If this is right

  • The framework produces strong results on the 32k- and 64k-token variants of ShopTrajQA.
  • Performance generalizes beyond shopping to other complex reasoning tasks.
  • Trajectories longer than any fixed context window become usable without truncation.
  • External storage plus tool calls replace the need to fit entire histories inside the model prompt.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same external-file-plus-code-tool pattern could be applied to other long sequential domains such as medical visit histories or financial transaction logs.
  • Success would reduce pressure to train ever-larger context windows by shifting the burden to reliable tool use.
  • The RLVR training loop may scale to agents that manage datasets orders of magnitude larger than current context limits allow.

Load-bearing premise

The premise that an agent can be trained to reliably formulate and execute code queries that retrieve the correct trajectory segments without introducing retrieval or execution errors.

What would settle it

A controlled test on ShopTrajQA where the agent receives queries requiring multi-table joins or time-range filters and either generates invalid SQL or yields lower accuracy than a baseline LLM given the full trajectory in context.

Figures

Figures reproduced from arXiv: 2606.07995 by Anurag Kashyap, Besnik Fetahu, Bing Yin, Hejie Cui, Hongye Liu, Ricardo Henao, Rongmei Lin.

Figure 1
Figure 1. Figure 1: Comparison between ShopTrajQA and OPeRA. The first row illustrates the trajectory generation pipeline [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Retrieval QA Construction: the model generates and verifies executable QA from shopping trajectory [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Customer-Agent framework. Red text denotes shopping trajectory information. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Number of action distribution for 32k and 64k [PITH_FULL_IMAGE:figures/full_fig_p013_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Token length distribution plot for 32k and 64k [PITH_FULL_IMAGE:figures/full_fig_p013_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: action token scatter for 32k and 64k setting. [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
read the original abstract

Understanding customer shopping trajectories is essential for enabling personalized shopping experiences. However, shopping records (i.e., customer's search, clicks, purchases, etc.) often span long time horizons over multiple years, resulting in extremely long trajectories that pose significant challenges for existing large language models (LLMs). Despite the importance of this problem, existing benchmarks are limited to short customer trajectories, while real-world trajectories from large e-commerce platforms are rarely accessible due to data privacy constraints. To address this gap, we introduce ShopTrajQA, a long-context evaluation benchmark constructed from real-world product information and simulated shopping trajectories. The dataset includes variants of up to 32k and 64k tokens, enabling systematic evaluation of model robustness under varying context lengths. Through comprehensive benchmarking of frontier LLMs, we identify critical performance gaps in reasoning over long shopping trajectory data. To address these challenges, we propose a Customer Agent Framework for ultra-long context management. Leveraging a Reinforcement Learning with Verifiable Rewards (RLVR) agentic training paradigm, our approach stores trajectories as external local files and trains the agent to autonomously retrieve and parse them through code-interpreter interactions (e.g., SQL queries), effectively bypassing the fixed in-context window constraints of LLMs. Experimental results demonstrate that our framework achieves strong performance for ShopTrajQA and shows generalization to other complex reasoning tasks.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper introduces ShopTrajQA, a benchmark constructed from real-world product data and simulated shopping trajectories with variants up to 32k and 64k tokens, benchmarks frontier LLMs to identify reasoning gaps on long contexts, and proposes the Customer Agent Framework. This framework uses an RLVR-trained agent to store trajectories in external files and autonomously retrieve/parse them via code-interpreter interactions (e.g., SQL queries), claiming this bypasses LLM context-window limits and yields strong ShopTrajQA performance plus generalization to other complex reasoning tasks.

Significance. If the empirical results hold, the work could meaningfully advance practical long-context handling in e-commerce personalization by demonstrating tool-augmented external access as a scalable alternative to in-context processing. The introduction of a privacy-respecting, real-world-derived benchmark is a positive contribution, but the complete absence of any quantitative metrics, ablations, or implementation details in the manuscript prevents assessment of whether the claimed bypass and generalization are achieved.

major comments (2)
  1. [Abstract] Abstract: the central claim of 'strong performance' and effective bypassing of context limits via the RLVR code-interpreter agent is unsupported because the manuscript supplies no metrics, ablation results, error bars, success rates on 64k-token variants, or details on reward density for query failures.
  2. [Abstract] Abstract: the assumption that the agent reliably discovers schema/query logic for multi-year trajectories and that the code interpreter handles arbitrarily long files without truncation is stated but not demonstrated; no evidence is given that RLVR training produces the required autonomous retrieval behavior on the longest variants.
minor comments (2)
  1. [Abstract] The abstract refers to 'comprehensive benchmarking of frontier LLMs' and 'experimental results' but provides none of the actual numbers, model names, or tables that would allow readers to evaluate the performance gaps or the framework's gains.
  2. Notation for the Customer Agent Framework and RLVR paradigm is introduced without definitions or pseudocode, making the training loop difficult to reconstruct.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We appreciate the referee's careful reading and the identification of areas where additional empirical support is needed. We will revise the manuscript accordingly to strengthen the presentation of our results.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim of 'strong performance' and effective bypassing of context limits via the RLVR code-interpreter agent is unsupported because the manuscript supplies no metrics, ablation results, error bars, success rates on 64k-token variants, or details on reward density for query failures.

    Authors: The referee is correct that the submitted manuscript does not include these specific quantitative details. We will add a comprehensive experimental results section with tables reporting success rates on ShopTrajQA variants up to 64k tokens, ablation studies comparing RLVR to baselines, error bars from repeated experiments, and details on the reward design including handling of query failures. This revision will directly support the claims made in the abstract. revision: yes

  2. Referee: [Abstract] Abstract: the assumption that the agent reliably discovers schema/query logic for multi-year trajectories and that the code interpreter handles arbitrarily long files without truncation is stated but not demonstrated; no evidence is given that RLVR training produces the required autonomous retrieval behavior on the longest variants.

    Authors: We agree that the current manuscript states these capabilities without providing direct evidence or demonstrations for the longest trajectories. In the revised version, we will include implementation details on the RLVR training process, examples of schema discovery and query generation, confirmation that the code interpreter processes files up to the required lengths without truncation, and experimental results demonstrating the autonomous retrieval behavior on 64k-token variants. We will also discuss any limitations observed. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical framework and benchmark with independent evaluation

full rationale

The paper introduces a new benchmark (ShopTrajQA) and an agent framework that stores trajectories externally and trains via RLVR for code-interpreter retrieval. No equations, fitted parameters, or self-referential definitions appear in the provided text. Performance claims are presented as experimental outcomes on the benchmark rather than derivations that reduce to inputs by construction. No load-bearing self-citations or uniqueness theorems are invoked. The approach is an engineering solution whose validity rests on external empirical results, not internal redefinition.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

The central claim rests on the unverified effectiveness of RLVR training for autonomous tool use and on the representativeness of the simulated trajectories; no independent evidence for either is supplied in the abstract.

axioms (1)
  • domain assumption Simulated shopping trajectories constructed from real product information accurately reflect real-world customer behavior patterns
    The benchmark is built from simulated trajectories; this assumption underpins all downstream claims about real-world utility.
invented entities (1)
  • Customer Agent Framework no independent evidence
    purpose: Managing ultra-long shopping trajectories by external storage and code-interpreter retrieval
    New agent architecture introduced in the paper; no independent evidence outside the abstract is provided.

pith-pipeline@v0.9.1-grok · 5796 in / 1322 out tokens · 25361 ms · 2026-06-27T20:03:30.677422+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

118 extracted references · 47 canonical work pages · 10 internal anchors

  1. [1]

    Advances in Neural Information Processing Systems , volume=

    Learning to reason with search for llms via reinforcement learning , author=. Advances in Neural Information Processing Systems , volume=

  2. [2]

    Nature , volume=

    DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning , author=. Nature , volume=. 2025 , publisher=

    2025

  3. [3]

    arXiv preprint arXiv:2211.12588 , year=

    Program of thoughts prompting: Disentangling computation from reasoning for numerical reasoning tasks , author=. arXiv preprint arXiv:2211.12588 , year=

    Pith/arXiv arXiv

  4. [4]

    arXiv preprint arXiv:2510.05381 , year=

    Context length alone hurts LLM performance despite perfect retrieval , author=. arXiv preprint arXiv:2510.05381 , year=

    arXiv

  5. [5]

    arXiv preprint arXiv:2504.11536 , year=

    Retool: Reinforcement learning for strategic tool use in llms , author=. arXiv preprint arXiv:2504.11536 , year=

    Pith/arXiv arXiv

  6. [6]

    International conference on machine learning , pages=

    Retrieval augmented language model pre-training , author=. International conference on machine learning , pages=. 2020 , organization=

    2020

  7. [7]

    Proceedings of the 28th International Conference on Computational Linguistics , pages=

    Constructing a multi-hop qa dataset for comprehensive evaluation of reasoning steps , author=. Proceedings of the 28th International Conference on Computational Linguistics , pages=

  8. [8]

    Proceedings of the 16th conference of the european chapter of the association for computational linguistics: main volume , pages=

    Leveraging passage retrieval with generative models for open domain question answering , author=. Proceedings of the 16th conference of the european chapter of the association for computational linguistics: main volume , pages=

  9. [9]

    arXiv preprint arXiv:2503.09516 , year=

    Search-r1: Training llms to reason and leverage search engines with reinforcement learning , author=. arXiv preprint arXiv:2503.09516 , year=

    Pith/arXiv arXiv

  10. [10]

    Advances in Neural Information Processing Systems , volume=

    Babilong: Testing the limits of llms with long context reasoning-in-a-haystack , author=. Advances in Neural Information Processing Systems , volume=

  11. [11]

    Advances in neural information processing systems , volume=

    Retrieval-augmented generation for knowledge-intensive nlp tasks , author=. Advances in neural information processing systems , volume=

  12. [12]

    arXiv preprint arXiv:2506.09820 , year=

    Cort: Code-integrated reasoning within thinking , author=. arXiv preprint arXiv:2506.09820 , year=

    arXiv

  13. [13]

    Transactions of the association for computational linguistics , volume=

    Lost in the middle: How language models use long contexts , author=. Transactions of the association for computational linguistics , volume=

  14. [14]

    Advances in neural information processing systems , volume=

    Training language models to follow instructions with human feedback , author=. Advances in neural information processing systems , volume=

  15. [15]

    Findings of the Association for Computational Linguistics: EMNLP 2023 , pages=

    Measuring and narrowing the compositionality gap in language models , author=. Findings of the Association for Computational Linguistics: EMNLP 2023 , pages=

    2023

  16. [16]

    2025 , url =

    Claude Haiku 4.5 , author =. 2025 , url =

    2025

  17. [17]

    2025 , url =

    Introducing GPT-OSS , author =. 2025 , url =

    2025

  18. [18]

    arXiv preprint arXiv:2206.06588 , year=

    Shopping queries dataset: A large-scale ESCI benchmark for improving product search , author=. arXiv preprint arXiv:2206.06588 , year=

    arXiv

  19. [19]

    arXiv preprint arXiv:2503.05592 , year=

    R1-searcher: Incentivizing the search capability in llms via reinforcement learning , author=. arXiv preprint arXiv:2503.05592 , year=

    Pith/arXiv arXiv

  20. [20]

    Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval , pages=

    Empowering large language models: Tool learning for real-world interaction , author=. Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval , pages=

  21. [21]

    arXiv preprint arXiv:2510.19363 , year=

    Loongrl: Reinforcement learning for advanced reasoning over long contexts , author=. arXiv preprint arXiv:2510.19363 , year=

    arXiv

  22. [22]

    arXiv preprint arXiv:2506.05606 , year=

    Opera: A dataset of observation, persona, rationale, and action for evaluating llms on human online shopping behavior simulation , author=. arXiv preprint arXiv:2506.05606 , year=

    Pith/arXiv arXiv

  23. [23]

    arXiv preprint arXiv:2510.07230 , year=

    Customer-R1: Personalized simulation of human behaviors via RL-based LLM agent in online shopping , author=. arXiv preprint arXiv:2510.07230 , year=

    arXiv

  24. [24]

    arXiv preprint arXiv:2505.09388 , year=

    Qwen3 technical report , author=. arXiv preprint arXiv:2505.09388 , year=

    Pith/arXiv arXiv

  25. [25]

    Proceedings of the 2018 conference on empirical methods in natural language processing , pages=

    HotpotQA: A dataset for diverse, explainable multi-hop question answering , author=. Proceedings of the 2018 conference on empirical methods in natural language processing , pages=

    2018

  26. [26]

    Advances in Neural Information Processing Systems , volume=

    Dapo: An open-source llm reinforcement learning system at scale , author=. Advances in Neural Information Processing Systems , volume=

  27. [27]

    arXiv preprint arXiv:2507.17842 , year=

    Shop-r1: Rewarding llms to simulate human behavior in online shopping via reinforcement learning , author=. arXiv preprint arXiv:2507.17842 , year=

    arXiv

  28. [28]

    arXiv preprint arXiv:2509.01055 , year=

    Verltool: Towards holistic agentic reinforcement learning with tool use , author=. arXiv preprint arXiv:2509.01055 , year=

    arXiv

  29. [29]

    arXiv preprint arXiv:2510.10649 , year=

    Unlocking exploration in rlvr: Uncertainty-aware advantage shaping for deeper reasoning , author=. arXiv preprint arXiv:2510.10649 , year=

    Pith/arXiv arXiv

  30. [30]

    arXiv preprint arXiv:2604.10734 , year=

    Self-correcting rag: Enhancing faithfulness via mmkp context selection and nli-guided mcts , author=. arXiv preprint arXiv:2604.10734 , year=

    Pith/arXiv arXiv

  31. [31]

    2026 , eprint=

    Semantic-Aware Logical Reasoning via a Semiotic Framework , author=. 2026 , eprint=

    2026

  32. [32]

    2026 , eprint=

    Logical Phase Transitions: Understanding Collapse in LLM Logical Reasoning , author=. 2026 , eprint=

    2026

  33. [33]

    arXiv preprint arXiv:2604.05516 , year=

    Coupling Macro Dynamics and Micro States for Long-Horizon Social Simulation , author=. arXiv preprint arXiv:2604.05516 , year=

    Pith/arXiv arXiv

  34. [34]

    2026 , eprint=

    STRIDE-ED: A Strategy-Grounded Stepwise Reasoning Framework for Empathetic Dialogue Systems , author=. 2026 , eprint=

    2026

  35. [35]

    arXiv preprint arXiv:2512.06690 , year=

    Think-While-Generating: On-the-Fly Reasoning for Personalized Long-Form Generation , author=. arXiv preprint arXiv:2512.06690 , year=

    arXiv

  36. [36]

    2026 , eprint=

    GCoT-Decoding: Unlocking Deep Reasoning Paths for Universal Question Answering , author=. 2026 , eprint=

    2026

  37. [37]

    2026 , eprint=

    DimMem: Dimensional Structuring for Efficient Long-Term Agent Memory , author=. 2026 , eprint=

    2026

  38. [38]

    Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing , pages=

    F ^2 Bench: An Open-ended Fairness Evaluation Benchmark for LLMs with Factuality Considerations , author=. Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing , pages=

    2025

  39. [39]

    Findings of the Association for Computational Linguistics: ACL 2025 , pages=

    McBE: A Multi-task Chinese Bias Evaluation Benchmark for Large Language Models , author=. Findings of the Association for Computational Linguistics: ACL 2025 , pages=

    2025

  40. [40]

    arXiv preprint arXiv:2604.10101 , year=

    Who Wrote This Line? Evaluating the Detection of LLM-Generated Classical Chinese Poetry , author=. arXiv preprint arXiv:2604.10101 , year=

    Pith/arXiv arXiv

  41. [41]

    2026 , eprint=

    ANCHOR: Abductive Network Construction with Hierarchical Orchestration for Reliable Probability Inference in Large Language Models , author=. 2026 , eprint=

    2026

  42. [42]

    arXiv preprint arXiv:2603.11863 , year=

    CreativeBench: Benchmarking and Enhancing Machine Creativity via Self-Evolving Challenges , author=. arXiv preprint arXiv:2603.11863 , year=

    Pith/arXiv arXiv

  43. [43]

    arXiv preprint arXiv:2604.07165 , year=

    Reason in chains, learn in trees: Self-rectification and grafting for multi-turn agent policy optimization , author=. arXiv preprint arXiv:2604.07165 , year=

    Pith/arXiv arXiv

  44. [44]

    arXiv preprint arXiv:2603.16060 , year=

    Arise: Agent reasoning with intrinsic skill evolution in hierarchical reinforcement learning , author=. arXiv preprint arXiv:2603.16060 , year=

    arXiv

  45. [45]

    DTCRS : Dynamic Tree Construction for Recursive Summarization

    Luo, Guanran and Jian, Zhongquan and Qiu, Wentao and Wang, Meihong and Wu, Qingqiang. DTCRS : Dynamic Tree Construction for Recursive Summarization. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025. doi:10.18653/v1/2025.acl-long.536

    work page doi:10.18653/v1/2025.acl-long.536 2025

  46. [46]

    Companion of the 2024 International Conference on Management of Data,

    Changlong Yu and Xin Liu and Jefferson Maia and Yang Li and Tianyu Cao and Yifan Gao and Yangqiu Song and Rahul Goutam and Haiyang Zhang and Bing Yin and Zheng Li , editor =. Companion of the 2024 International Conference on Management of Data,. 2024 , url =. doi:10.1145/3626246.3653398 , timestamp =

    work page doi:10.1145/3626246.3653398 2024

  47. [47]

    FolkScope: Intention Knowledge Graph Construction for E-commerce Commonsense Discovery , booktitle =

    Changlong Yu and Weiqi Wang and Xin Liu and Jiaxin Bai and Yangqiu Song and Zheng Li and Yifan Gao and Tianyu Cao and Bing Yin , editor =. FolkScope: Intention Knowledge Graph Construction for E-commerce Commonsense Discovery , booktitle =. 2023 , url =. doi:10.18653/V1/2023.FINDINGS-ACL.76 , timestamp =

    work page doi:10.18653/v1/2023.findings-acl.76 2023

  48. [48]

    Tom B. Brown and Benjamin Mann and Nick Ryder and Melanie Subbiah and Jared Kaplan and Prafulla Dhariwal and Arvind Neelakantan and Pranav Shyam and Girish Sastry and Amanda Askell and Sandhini Agarwal and Ariel Herbert. Language Models are Few-Shot Learners , booktitle =. 2020 , url =

    2020

  49. [49]

    , author=

    Measuring nominal scale agreement among many raters. , author=. Psychological bulletin , volume=. 1971 , publisher=

    1971

  50. [50]

    SayCanPay: Heuristic Planning with Large Language Models Using Learnable Domain Knowledge , booktitle =

    Rishi Hazra and Pedro Zuidberg Dos Martires and Luc De Raedt , editor =. SayCanPay: Heuristic Planning with Large Language Models Using Learnable Domain Knowledge , booktitle =. 2024 , url =. doi:10.1609/AAAI.V38I18.29991 , timestamp =

    work page doi:10.1609/aaai.v38i18.29991 2024

  51. [51]

    Reflexion: language agents with verbal reinforcement learning , booktitle =

    Noah Shinn and Federico Cassano and Ashwin Gopinath and Karthik Narasimhan and Shunyu Yao , editor =. Reflexion: language agents with verbal reinforcement learning , booktitle =. 2023 , url =

    2023

  52. [52]

    The Llama 3 Herd of Models

    Abhimanyu Dubey and Abhinav Jauhri and Abhinav Pandey and Abhishek Kadian and Ahmad Al. The Llama 3 Herd of Models , journal =. 2024 , url =. doi:10.48550/ARXIV.2407.21783 , eprinttype =. 2407.21783 , timestamp =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2407.21783 2024

  53. [53]

    Detecting online commercial intention

    Honghua (Kathy) Dai and Lingzhi Zhao and Zaiqing Nie and Ji. Detecting online commercial intention. Proceedings of the 15th international conference on World Wide Web,. 2006 , url =. doi:10.1145/1135777.1135902 , timestamp =

    work page doi:10.1145/1135777.1135902 2006

  54. [54]

    Yu , editor =

    Chenwei Zhang and Wei Fan and Nan Du and Philip S. Yu , editor =. Mining User Intentions from Medical Queries:. Proceedings of the 25th International Conference on World Wide Web,. 2016 , url =. doi:10.1145/2872427.2874810 , timestamp =

    work page doi:10.1145/2872427.2874810 2016

  55. [55]

    McAuley , editor =

    Jianmo Ni and Jiacheng Li and Julian J. McAuley , editor =. Justifying Recommendations using Distantly-Labeled Reviews and Fine-Grained Aspects , booktitle =. 2019 , url =. doi:10.18653/V1/D19-1018 , timestamp =

    work page doi:10.18653/v1/d19-1018 2019

  56. [56]

    Susan Zhang and Stephen Roller and Naman Goyal and Mikel Artetxe and Moya Chen and Shuohui Chen and Christopher Dewan and Mona T. Diab and Xian Li and Xi Victoria Lin and Todor Mihaylov and Myle Ott and Sam Shleifer and Kurt Shuster and Daniel Simig and Punit Singh Koura and Anjali Sridhar and Tianlu Wang and Luke Zettlemoyer , title =. CoRR , volume =. 2...

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2205.01068 2022

  57. [57]

    Yu and Xue Wang and Jian Wang , editor =

    Zhenyun Hao and Jianing Hao and Zhaohui Peng and Senzhang Wang and Philip S. Yu and Xue Wang and Jian Wang , editor =. Dy-HIEN: Dynamic Evolution based Deep Hierarchical Intention Network for Membership Prediction , booktitle =. 2022 , url =. doi:10.1145/3488560.3498517 , timestamp =

    work page doi:10.1145/3488560.3498517 2022

  58. [58]

    2000 , publisher=

    Intention , author=. 2000 , publisher=

    2000

  59. [59]

    Bridging Language and Items for Retrieval and Recommendation: Benchmarking LLMs as Semantic Encoders

    Yupeng Hou and Jiacheng Li and Zhankui He and An Yan and Xiusi Chen and Julian J. McAuley , title =. CoRR , volume =. 2024 , url =. doi:10.48550/ARXIV.2403.03952 , eprinttype =. 2403.03952 , timestamp =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2403.03952 2024

  60. [60]

    Journal of memory and language , volume=

    The representation of scripts in memory , author=. Journal of memory and language , volume=. 1985 , publisher=

    1985

  61. [61]

    Multimedia Generative Script Learning for Task Planning , booktitle =

    Qingyun Wang and Manling Li and Hou Pong Chan and Lifu Huang and Julia Hockenmaier and Girish Chowdhary and Heng Ji , editor =. Multimedia Generative Script Learning for Task Planning , booktitle =. 2023 , url =. doi:10.18653/V1/2023.FINDINGS-ACL.63 , timestamp =

    work page doi:10.18653/v1/2023.findings-acl.63 2023

  62. [62]

    IJCAI , volume=

    Scripts, plans, and knowledge , author=. IJCAI , volume=. 1975 , organization=

    1975

  63. [63]

    Cognitive psychology , volume=

    Scripts in memory for text , author=. Cognitive psychology , volume=. 1979 , publisher=

    1979

  64. [64]

    Llama 2: Open Foundation and Fine-Tuned Chat Models

    Hugo Touvron and Louis Martin and Kevin Stone and Peter Albert and Amjad Almahairi and Yasmine Babaei and Nikolay Bashlykov and Soumya Batra and Prajjwal Bhargava and Shruti Bhosale and Dan Bikel and Lukas Blecher and Cristian Canton. Llama 2: Open Foundation and Fine-Tuned Chat Models , journal =. 2023 , url =. doi:10.48550/ARXIV.2307.09288 , eprinttype ...

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2307.09288 2023

  65. [65]

    CoRR , volume =

    Yinhan Liu and Myle Ott and Naman Goyal and Jingfei Du and Mandar Joshi and Danqi Chen and Omer Levy and Mike Lewis and Luke Zettlemoyer and Veselin Stoyanov , title =. CoRR , volume =. 2019 , url =. 1907.11692 , timestamp =

    Pith/arXiv arXiv 2019

  66. [66]

    The Eleventh International Conference on Learning Representations,

    Pengcheng He and Jianfeng Gao and Weizhu Chen , title =. The Eleventh International Conference on Learning Representations,. 2023 , url =

    2023

  67. [67]

    Findings of the Association for Computational Linguistics:

    Weiqi Wang and Tianqing Fang and Wenxuan Ding and Baixuan Xu and Xin Liu and Yangqiu Song and Antoine Bosselut , editor =. Findings of the Association for Computational Linguistics:. 2023 , url =. doi:10.18653/V1/2023.FINDINGS-EMNLP.902 , timestamp =

    work page doi:10.18653/v1/2023.findings-emnlp.902 2023

  68. [68]

    Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers),

    Weiqi Wang and Tianqing Fang and Chunyang Li and Haochen Shi and Wenxuan Ding and Baixuan Xu and Zhaowei Wang and Jiaxin Bai and Xin Liu and Cheng Jiayang and Chunkit Chan and Yangqiu Song , editor =. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers),. 2024 , url =

    2024

  69. [69]

    Smith and Yejin Choi and Hannaneh Hajishirzi , editor =

    Jiacheng Liu and Wenya Wang and Dianzhuo Wang and Noah A. Smith and Yejin Choi and Hannaneh Hajishirzi , editor =. Vera:. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing,. 2023 , url =. doi:10.18653/V1/2023.EMNLP-MAIN.81 , timestamp =

    work page doi:10.18653/v1/2023.emnlp-main.81 2023

  70. [70]

    Gemma: Open Models Based on Gemini Research and Technology

    Thomas Mesnard and Cassidy Hardin and Robert Dadashi and Surya Bhupatiraju and Shreya Pathak and Laurent Sifre and Morgane Rivi. Gemma: Open Models Based on Gemini Research and Technology , journal =. 2024 , url =. doi:10.48550/ARXIV.2403.08295 , eprinttype =. 2403.08295 , timestamp =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2403.08295 2024

  71. [71]

    Kingma and Jimmy Ba , editor =

    Diederik P. Kingma and Jimmy Ba , editor =. Adam:. 3rd International Conference on Learning Representations,. 2015 , url =

    2015

  72. [72]

    2021 , url =

    Tianqing Fang and Hongming Zhang and Weiqi Wang and Yangqiu Song and Bin He , editor =. 2021 , url =. doi:10.1145/3442381.3450117 , timestamp =

    work page doi:10.1145/3442381.3450117 2021

  73. [73]

    Benchmarking Commonsense Knowledge Base Population with an Effective Evaluation Dataset , booktitle =

    Tianqing Fang and Weiqi Wang and Sehyun Choi and Shibo Hao and Hongming Zhang and Yangqiu Song and Bin He , editor =. Benchmarking Commonsense Knowledge Base Population with an Effective Evaluation Dataset , booktitle =. 2021 , url =. doi:10.18653/V1/2021.EMNLP-MAIN.705 , timestamp =

    work page doi:10.18653/v1/2021.emnlp-main.705 2021

  74. [74]

    Liu , title =

    Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu , title =. J. Mach. Learn. Res. , volume =. 2020 , url =

    2020

  75. [75]

    Kelvin J. L. Koa and Yunshan Ma and Ritchie Ng and Tat. Learning to Generate Explainable Stock Predictions using Self-Reflective Large Language Models , booktitle =. 2024 , url =. doi:10.1145/3589334.3645611 , timestamp =

    work page doi:10.1145/3589334.3645611 2024

  76. [76]

    TasTe: Teaching Large Language Models to Translate through Self-Reflection , booktitle =

    Yutong Wang and Jiali Zeng and Xuebo Liu and Fandong Meng and Jie Zhou and Min Zhang , editor =. TasTe: Teaching Large Language Models to Translate through Self-Reflection , booktitle =. 2024 , url =. doi:10.18653/V1/2024.ACL-LONG.333 , timestamp =

    work page doi:10.18653/v1/2024.acl-long.333 2024

  77. [77]

    Hwang and Chandra Bhagavatula and Ronan Le Bras and Jeff Da and Keisuke Sakaguchi and Antoine Bosselut and Yejin Choi , title =

    Jena D. Hwang and Chandra Bhagavatula and Ronan Le Bras and Jeff Da and Keisuke Sakaguchi and Antoine Bosselut and Yejin Choi , title =. Thirty-Fifth. 2021 , url =. doi:10.1609/AAAI.V35I7.16792 , timestamp =

    work page doi:10.1609/aaai.v35i7.16792 2021

  78. [78]

    Smith and Yejin Choi , title =

    Maarten Sap and Ronan Le Bras and Emily Allaway and Chandra Bhagavatula and Nicholas Lourie and Hannah Rashkin and Brendan Roof and Noah A. Smith and Yejin Choi , title =. The Thirty-Third. 2019 , url =. doi:10.1609/AAAI.V33I01.33013027 , timestamp =

    work page doi:10.1609/aaai.v33i01.33013027 2019

  79. [79]

    Transformers: State-of-the-Art Natural Language Processing , booktitle =

    Thomas Wolf and Lysandre Debut and Victor Sanh and Julien Chaumond and Clement Delangue and Anthony Moi and Pierric Cistac and Tim Rault and R. Transformers: State-of-the-Art Natural Language Processing , booktitle =. 2020 , url =. doi:10.18653/V1/2020.EMNLP-DEMOS.6 , timestamp =

    work page doi:10.18653/v1/2020.emnlp-demos.6 2020

  80. [80]

    Reimers, I

    Nils Reimers and Iryna Gurevych , editor =. Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks , booktitle =. 2019 , url =. doi:10.18653/V1/D19-1410 , timestamp =

    work page doi:10.18653/v1/d19-1410 2019

Showing first 80 references.