Pith. sign in

REVIEW 6 major objections 6 minor 49 references

Aspect-Based Summarization with Self-Aspect Retrieval Enhanced Generation

T0 review · 6 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read Pruning a document to aspect-relevant sentences before LLM summarization improves the resulting summaries and leaves room for in-context examples.

desk verdict A sensible retrieval-pruning pipeline for aspect-based summarization that mostly supports its quality claims, but the token-limit headline is untested because the ICL evaluation never exercises long documents. read the letter →

arxiv 2504.13054 v1 pith:4GUG4HQC submitted 2025-04-17 cs.CL cs.AI

Yichao Feng , Shuai Zhao , Yueqiu Li , Luwei Xiao , Xiaobao Wu , Anh Tuan Luu This is my paper · ORCID
classification cs.CLcs.AI
keywords aspect-basedsummarizationlargelanguagemodelsin-contextlearningretrieval-enhancedgenerationdenseretrievaltokenlimitstextpruning
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 before asking an LLM to write an aspect-specific summary, you can improve both quality and efficiency by first retrieving the sentences most related to that aspect and pruning everything else. The proposed framework, SARESG, splits a long document into 256-word chunks, scores each sentence with an embedding model by cosine similarity to the aspect, keeps only the top-scoring sentences up to a word budget, and reorders them before generation. Across three benchmark datasets and three LLMs, the pruned-input summaries score higher on METEOR, ROUGE, and BERTScore than full-text prompting and than an existing selective-context baseline, and the token savings leave room for in-context examples. The paper argues that this reduces hallucination risk and lets smaller models approach larger zero-shot models.

What carries the argument

The load-bearing object is the self-aspect retrieval score of Eq. (2), $S_{i,j} = \cos(f(s_{i,j}), f(A))$, where $f$ is a pretrained embedding model, $s_{i,j}$ is a sentence inside a 256-word chunk, and $A$ is the aspect. That score drives the Top-W selection of Eq. (3): within each chunk the sentences are sorted by this score and kept until their cumulative word count reaches $W$, then the survivors are glued back in their original order. This mechanism converts a long-document summarization problem into a short-document one without retraining the LLM, and it is what frees token space for in-context learning.

What would settle it

Take a document with human-annotated sentences that appear in the gold aspect summary, run SARESG's pruning, and check how many gold-summary sentences fall below the Top-W cutoff; if a substantial fraction of the gold content is discarded while irrelevant sentences survive, the retrieval-ranking assumption is false. A cheaper version is an ablation that replaces the embedding scores with random sentence rankings: if random pruning performs as well as SARESG on the same metrics, the retrieval signal is not doing the work.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that relevance-based pruning of the input, driven by the aspect itself, is a sufficient alternative to either full-text prompting or length-based truncation for LLM aspect-based summarization. Rather than relying on prompt engineering or fitting the whole document into the context window, the system defines an aspect-aware selection step: embed each sentence and the aspect, rank sentences by cosine similarity, select the highest-scoring sentences per chunk until a word threshold is met, and concatenate the survivors in their original order. The experiments report consistent gains across Mistral and two Llama3 sizes, and the gains persist when an in-context example is added. The paper also claims the pruned input leaves token budget for such examples, which is why SARESG_ICL often beats truncated ICL.

Load-bearing premise

The assumption that carries the whole method is that an embedding model's cosine similarity between a sentence and the aspect phrase ranks sentences by how much they matter for the final summary, so pruning to the top-scoring sentences never throws away content the summary needs.

Editorial extensions

If this is right

  • If the retrieval-based pruning claim holds, long documents can be summarized by LLMs with small context windows without manually engineered prompts: the aspect name itself supplies the retrieval query.
  • Token savings from pruning allow one-shot in-context examples, and the tables suggest that an 8-billion-parameter model with SARESG_ICL can match a 70-billion-parameter zero-shot model on several metrics.
  • Because irrelevant sentences are removed before generation, the input the LLM sees is shorter and more topic-concentrated, which the paper argues lowers the chance of hallucinated or off-aspect content.
  • The chunk-size finding (256 words optimal in their test) implies that the method needs a coarse-enough context unit to judge relevance, and that sentence-level retrieval alone is weaker.

Reading between the lines

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

  • The paper does not directly measure hallucination; I infer that the claimed hallucination reduction is a plausible consequence of a shorter, focused input, but it needs a factuality-oriented evaluation to be established.
  • Because the pruning threshold $W$ and chunk size are dataset-dependent, the framework's practical value depends on having a cheap way to tune them per corpus; the paper's own limitations say the best pruning parameters vary with article length.
  • The embedding model's ranking quality is the hidden bottleneck; any improvement in sentence-aspect similarity, such as learned relevance models or aspect-aware query expansion, should transfer directly to better summaries.
  • A natural extension is to make retrieval iterative or hierarchical, first pruning at chunk level and then at sentence level, which could handle even longer inputs than the current single-pass scheme.
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

6 major / 6 minor

Summary. The paper proposes SARESG, a retrieve-and-prune framework for aspect-based summarization with large language models. The input document is split into fixed-size chunks (256 words), each sentence is scored by cosine similarity to the aspect embedding, the top-scoring sentences are kept up to an unspecified word threshold W, and the pruned document is fed to an LLM, optionally with one in-context learning (ICL) demonstration. Experiments on Ma-news, OAsum, and USB with Mistral (Mixtral) 8x7b, Llama3-8b, and Llama3-70b compare SARESG and SARESG_ICL against Original, Selective Context, and Truncated_ICL baselines using METEOR, ROUGE, and BERTScore. The paper claims that the framework achieves superior summarization quality and effectively mitigates the token-limitation problem, and it includes ablations on chunk size and on sentence- versus chunk-level retrieval.

Significance. If the reported results hold with proper controls, SARESG is a simple and practical contribution to aspect-based summarization: it offers a lightweight retrieval-based pruning mechanism that can free token space for ICL while improving aspect alignment. The paper's strengths are its breadth of models and datasets, the inclusion of a context-compression baseline (Selective Context), and the ablation studies on chunk size and retrieval granularity. However, the central token-mitigation claim is not experimentally tested, the pruning hyperparameters are undisclosed, and the quality comparisons lack statistical rigor; the paper's value is therefore conditional on additional controlled experiments.

major comments (6)
  1. [Section IV.A, Tables I-III, Abstract] The claim that the framework 'effectively mitigates the token limitation problem' is not supported by the reported experiments. Section IV.A states that 'due to the token limit in the ICL setting... we only retain samples with a length of less than 1024.' Consequently, every ICL comparison in Tables I-III is performed on documents that already fit comfortably within the 4096-token budget, and the retrieval-based pruning is never tested under the token pressure it is designed to relieve. No table or figure reports the actual number of input tokens consumed by SARESG_ICL versus Truncated_ICL. Please add long-document experiments (or at least report token counts and the distribution of input lengths) to substantiate the token-usage claim.
  2. [Section III.A, Eq. (3); Section IV.A] The pruning threshold W in Eq. (3) is never specified, and Section IV.A only says that input lengths were 'adjusted to ensure consistency with the baseline' without giving the resulting lengths. Without W or token counts, the gains of SARESG over Truncated_ICL could be due to a different effective context length rather than to better content selection. The Limitations section also acknowledges that 'the best pruning parameters will be varied as the length of articles varied.' Please report W, the resulting input lengths for each dataset, and a sensitivity analysis over W.
  3. [Table III (OAsum rows); Section V.D] The OAsum ICL failure is explained post hoc without reporting corrected numbers. In Table III, Truncated_ICL and SARESG_ICL score below the zero-shot Original and SARESG (e.g., METEOR 24.18 and 23.72 vs 24.34 and 25.38). Section V.D attributes this to an unrepresentative random ICL sample and says additional experiments with different samples were conducted, but the corrected results do not appear in any table. This leaves the main results table inconsistent with the text and prevents the reader from assessing the ICL claim for OAsum. Please report the corrected OAsum ICL numbers (or a separate table) and state the sample-selection protocol before the experiments.
  4. [Tables I-III; Section IV.A] Tables I-III list methods 1, 2, 3, 5, and 6 but not row 4; if a fourth baseline was intended (e.g., Selective Context with ICL), it should be included for a complete comparison. More importantly, no error bars, variance, or significance tests are reported on any metric. Differences such as Ma-news METEOR 24.07 (SARESG) vs 22.91 (Original) in Table I may be within sampling variation; please add standard errors or paired significance tests across the evaluation samples.
  5. [Section I (Introduction); Section III] The Introduction states that the retrieval process is 'recursive, continuing until the text is pruned to the desired length,' but Section III describes only a single pass: chunk, score, select Top-W, concatenate. No recursive algorithm, termination criterion, or ablation is presented. Please either remove the word 'recursive' from the Introduction or specify and evaluate the recursive procedure.
  6. [Section V.C; Section IV.A] The chunk size of 256 appears to be selected based on ROUGE scores on the Ma-news evaluation set, with no indication of a held-out validation split. Since chunk size is a hyperparameter of SARESG, tuning it on the test set risks inflating the reported gains over baselines. Please describe how hyperparameters (chunk size, W) were chosen, ideally on a validation split.
minor comments (6)
  1. [Section IV.A] The sentence 'the jasper model on top of the stella A model called stella en 1.5b' is unclear; please give the exact model name and version used for retrieval.
  2. [Table IV caption] The caption contains a typo: 'Retrival' should be 'Retrieval', and the second row should be 'Chunk Retrieval' rather than 'Chunk Retrival'.
  3. [References] References [18] and [40] are the same paper (Frermann and Klementiev, ACL 2019); please merge them into a single citation.
  4. [Figure 1] The figure caption contains a stray space in 'V onderplanitz's', and the diagram text appears garbled; please regenerate the figure for clarity.
  5. [Section IV.A] The model name 'Mistral 8×7b' should be 'Mixtral 8x7b', and the truncation baseline is inconsistently named 'Truncated' and 'Truncated_ICL' across the text and tables.
  6. [Section IV.A] The filtering criterion 'we only retain samples with a length of less than 1024' is not tied to the earlier statement that the first 2000 rows of Ma-news and OAsum were selected; please report the resulting dataset sizes and the length statistics for each dataset.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the retrieval-pruning pipeline is an input selector, and all reported gains are measured against external baselines and metrics.

full rationale

The paper's central mechanism (Eqs. 2-3) selects and prunes input sentences by cosine similarity to the aspect embedding; it does not define or predict the reported summary-quality metrics. The summarized output is generated by an LLM and scored by METEOR, ROUGE, and BERTScore against reference summaries on the public Ma-news, OAsum, and USB benchmarks. No equation is fitted to these metrics, and no parameter is defined in terms of the target result. The comparisons to Original, Selective Context, and Truncated_ICL are external empirical baselines, so the quality claims are not forced by construction. Self-citations in the reference list are background citations and are not load-bearing for the method's derivation. The token-limitation claim has an evidentiary gap: the paper restricts ICL samples to under 1024 tokens and does not report token budgets for SARESG_ICL versus Truncated_ICL, so the headline mitigation claim is not directly demonstrated. However, lack of direct experimental support is a validity concern, not circularity under the stated rubric. The tuning of chunk size and the post-hoc swapping of an ICL sample after observing a performance drop are reproducibility and selection concerns, but they do not reduce any reported quantity to an input by definition.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the assumption that embedding similarity is a good proxy for summary relevance, plus several unstated hyperparameters (chunk size, W). No new entities are introduced.

free parameters (3)
  • chunk size = 256 words
    Set to 256 based on ablations on the Ma-news dataset (Section V-C); the paper states larger chunks retrieve too few, smaller chunks fragment.
  • pruning word threshold W = unspecified
    Defined in Eq. (3) as the cumulative word target for selected sentences, but its value used in experiments is never reported, making the pruning ratio unreproducible.
  • ICL sample selection = one random sample, later replaced for OAsum
    The one-shot demo is randomly chosen; when the OAsum sample turned out to be degenerate, the authors ran 'additional experiments' with different samples whose results are not included in the main tables.
assumptions (4)
  • domain assumption Cosine similarity between sentence and aspect embeddings is a valid relevance signal for summary content.
    Eq. (2) uses cosine similarity to rank sentences; if the embedding model fails to capture aspect relevance, pruning removes useful content.
  • domain assumption Pruning to W words preserves enough context for faithful summarization.
    The method assumes that selecting the top-W sentences by similarity leaves all information needed for an aspect-specific summary.
  • domain assumption The pretrained embedding model (stella en 1.5b with jasper) is suitable for the three datasets.
    No fine-tuning is performed; the retrieval quality is taken on faith from the pretrained model.
  • standard math Standard vector arithmetic (cosine similarity, top-k) is used.
    Equations (2)-(3) rely on standard operations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Aspect-Based Summarization with Self-Aspect Retrieval Enhanced Generation." pith.science (2026). https://pith.science/paper/4GUG4HQC

@misc{pith2026250413054,
  author       = {Pith},
  title        = {Pith review of: Aspect-Based Summarization with Self-Aspect Retrieval Enhanced Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4GUG4HQC}},
  note         = {Machine review of arXiv:2504.13054}
}
read the original abstract

Aspect-based summarization aims to generate summaries tailored to specific aspects, addressing the resource constraints and limited generalizability of traditional summarization approaches. Recently, large language models have shown promise in this task without the need for training. However, they rely excessively on prompt engineering and face token limits and hallucination challenges, especially with in-context learning. To address these challenges, in this paper, we propose a novel framework for aspect-based summarization: Self-Aspect Retrieval Enhanced Summary Generation. Rather than relying solely on in-context learning, given an aspect, we employ an embedding-driven retrieval mechanism to identify its relevant text segments. This approach extracts the pertinent content while avoiding unnecessary details, thereby mitigating the challenge of token limits. Moreover, our framework optimizes token usage by deleting unrelated parts of the text and ensuring that the model generates output strictly based on the given aspect. With extensive experiments on benchmark datasets, we demonstrate that our framework not only achieves superior performance but also effectively mitigates the token limitation problem.

Figures

Figures reproduced from arXiv: 2504.13054 by the authors.

Figure 1
Figure 1. This diagram illustrates the generation process of the system. The model first evaluates the overall length of the input text. If the text is shorter than a predefined threshold, it is directly fed into the LLM for summary generation. For longer texts, the system splits the input into smaller chunks, uses an embedding model to assign relevance scores to each chunk, filters out the least relevant parts, and reconstru… view at source ↗
Figure 2
Figure 2. Experiment results for three different chunk sizes on MA-news dataset. Results generated with Llama3-70b. However, for the OAsum dataset, we observed an opposite trend: the ICL scenarios, including both truncated and SARESG approaches, achieved lower scores compared to the zero-shot cases. Upon further investigation, we identified the primary reason for this discrepancy. The randomly chosen sample used for the ICL e… view at source ↗
Figure 3
Figure 3. Radar charts were created for all ICL samples tested on the MA-news dataset, with results generated using the Llama3-8b model. Blue is representing R1, orange is representing R2, and green is representing Rl. The captions represent the aspects of the samples, with six points on each chart corresponding to the aspects of the testing set. examine larger-scale datasets or develop methods to identify optimal samples for… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 26 canonical work pages

  1. [1]

    Senticvec: Toward robust and human-centric neurosymbolic sentiment analysis,

    X. Zhang, R. Mao, and E. Cambria, “Senticvec: Toward robust and human-centric neurosymbolic sentiment analysis,” ACL Findings, 2024

    work page 2024

  2. [2]

    Vanessa: Visual connotation and aesthetic attributes understanding network for multimodal aspect-based sentiment analysis,

    L. Xiao, R. Mao, X. Zhang, L. He, and E. Cambria, “Vanessa: Visual connotation and aesthetic attributes understanding network for multimodal aspect-based sentiment analysis,” in Findings of the Association for Computational Linguistics: EMNLP 2024 , 2024, pp. 11 486–11 500

    2024

  3. [3]

    Discovering the cognition behind language: Financial metaphor analysis with metapro,

    R. Mao, K. Du, Y . Ma, L. Zhu, and E. Cambria, “Discovering the cognition behind language: Financial metaphor analysis with metapro,” in 2023 ICDM. IEEE, 2023, pp. 1211–1216

    work page 2023

  4. [4]

    An evaluation of reasoning capabilities of large language models in financial sentiment analysis,

    K. Du, F. Xing, R. Mao, and E. Cambria, “An evaluation of reasoning capabilities of large language models in financial sentiment analysis,” in IEEE Conference on Artificial Intelligence (IEEE CAI), Singapore , 2024

    work page 2024

  5. [5]

    Llamandement: Large language models for summarization of french legislative proposals,

    J. Gesnouin, Y . Tannier, C. G. Da Silva, H. Tapory, C. Brier, H. Simon, R. Rozenberg, H. Woehrel, M. E. Yakaabi, T. Binder et al., “Llamandement: Large language models for summarization of french legislative proposals,” arXiv preprint arXiv:2401.16182 , 2024

    arXiv 2024

  6. [6]

    A sentence is known by the company it keeps: Improving legal document summarization using deep clustering,

    D. Jain, M. D. Borah, and A. Biswas, “A sentence is known by the company it keeps: Improving legal document summarization using deep clustering,” Artificial Intelligence and Law , vol. 32, no. 1, pp. 165–200, 2024

    work page 2024

  7. [7]

    The biases of pre-trained language models: An empirical study on prompt-based sentiment analysis and emotion detection,

    R. Mao, Q. Liu, K. He, W. Li, and E. Cambria, “The biases of pre-trained language models: An empirical study on prompt-based sentiment analysis and emotion detection,” IEEE transactions on affective computing, vol. 14, no. 3, pp. 1743–1753, 2022

    2022

  8. [8]

    Short text topic modeling with topic distribution quantization and negative sampling decoder,

    X. Wu, C. Li, Y . Zhu, and Y . Miao, “Short text topic modeling with topic distribution quantization and negative sampling decoder,” in Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), Nov. 2020

    work page 2020

Show all 49 references
  1. [9]

    Mitigating data sparsity for short text topic modeling by topic-semantic contrastive learning,

    X. Wu, A. T. Luu, and X. Dong, “Mitigating data sparsity for short text topic modeling by topic-semantic contrastive learning,” arXiv preprint arXiv:2211.12878, 2022

    2022 arXiv

  2. [10]

    Fastopic: A fast, adaptive, stable, and transferable topic modeling paradigm,

    X. Wu, T. Nguyen, D. C. Zhang, W. Y . Wang, and A. T. Luu, “Fastopic: A fast, adaptive, stable, and transferable topic modeling paradigm,” arXiv preprint arXiv:2405.17978, 2024

    2024 arXiv

  3. [11]

    A survey on neural topic models: methods, applications, and challenges,

    X. Wu, T. Nguyen, and A. T. Luu, “A survey on neural topic models: methods, applications, and challenges,” arXiv preprint arXiv:2401.15351, 2024

    2024 arXiv

  4. [12]

    Benchmarking large language models for news summarization,

    T. Zhang, F. Ladhak, E. Durmus, P. Liang, K. McKeown, and T. B. Hashimoto, “Benchmarking large language models for news summarization,” Transactions of the Association for Computational Linguistics, vol. 12, pp. 39–57, 2024

    2024

  5. [13]

    To repeat or not to repeat: Insights from scaling llm under token-crisis,

    F. Xue, Y . Fu, W. Zhou, Z. Zheng, and Y . You, “To repeat or not to repeat: Insights from scaling llm under token-crisis,” Advances in Neural Information Processing Systems , vol. 36, 2024

    2024

  6. [14]

    Longlora: Efficient fine-tuning of long-context large language models,

    Y . Chen, S. Qian, H. Tang, X. Lai, Z. Liu, S. Han, and J. Jia, “Longlora: Efficient fine-tuning of long-context large language models,” arXiv preprint arXiv:2309.12307, 2023

    2023 arXiv

  7. [15]

    Bench: Extending long context evaluation beyond 100k tokens,

    X. Zhang, Y . Chen, S. Hu, Z. Xu, J. Chen, M. Hao, X. Han, Z. Thai, S. Wang, Z. Liu et al., “Bench: Extending long context evaluation beyond 100k tokens,” in ACL, 2024, pp. 15 262–15 277

    2024

  8. [16]

    A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions,

    L. Huang, W. Yu, W. Ma, W. Zhong, Z. Feng, H. Wang, Q. Chen, W. Peng, X. Feng, B. Qin et al. , “A survey on hallucination in large language models: Principles, taxonomy, challenges, and open questions,” ACM Transactions on Information Systems , 2023

    2023

  9. [17]

    A survey on in-context learning,

    Q. Dong, L. Li, D. Dai, C. Zheng, J. Ma, R. Li, H. Xia, J. Xu, Z. Wu, B. Chang, X. Sun, L. Li, and Z. Sui, “A survey on in-context learning,” in Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, Y . Al-Onaizan, M. Bansal, and Y .-N. Chen, ...

    2024

  10. [18]

    Inducing document structure for aspect-based summarization,

    L. Frermann and A. Klementiev, “Inducing document structure for aspect-based summarization,” in Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics , 2019, pp. 6263–6273

    2019

  11. [19]

    Bridging towers of multi-task learning with a gating mechanism for aspect-based sentiment analysis and sequential metaphor identification,

    R. Mao and X. Li, “Bridging towers of multi-task learning with a gating mechanism for aspect-based sentiment analysis and sequential metaphor identification,” in Proceedings of the AAAI conference on artificial intelligence, vol. 35, no. 15, 2021, pp. 13 534–13 542

    2021

  12. [20]

    Atlantis: Aesthetic-oriented multiple granularities fusion network for joint multimodal aspect-based sentiment analysis,

    L. Xiao, X. Wu, J. Xu, W. Li, C. Jin, and L. He, “Atlantis: Aesthetic-oriented multiple granularities fusion network for joint multimodal aspect-based sentiment analysis,” Information Fusion, vol. 106, p. 102304, 2024

    2024

  13. [21]

    Lexabsumm: Aspect-based summarization of legal decisions,

    T. Santosh, M. Aly, and M. Grabmair, “Lexabsumm: Aspect-based summarization of legal decisions,” arXiv preprint arXiv:2404.00594 , 2024

    2024 arXiv

  14. [22]

    Aspectnews: Aspect-oriented summarization of news documents,

    O. Ahuja, J. Xu, A. Gupta, K. Horecka, and G. Durrett, “Aspectnews: Aspect-oriented summarization of news documents,” arXiv preprint arXiv:2110.08296, 2021

    2021 arXiv

  15. [23]

    Extractive summarization as text matching,

    M. Zhong, P. Liu, Y . Chen, D. Wang, X. Qiu, and X. Huang, “Extractive summarization as text matching,” in Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , D. Jurafsky, J. Chai, N. Schluter, and J. Tetreault, Eds., Jul. 2020, pp. 6197–6208

    2020

  16. [24]

    Longformer: The long-document transformer,

    I. Beltagy, M. E. Peters, and A. Cohan, “Longformer: The long-document transformer,” arXiv preprint arXiv:2004.05150 , 2020

    2004 arXiv

  17. [25]

    Feamix: Feature mix with memory batch based on self-consistency learning for code generation and code translation,

    S. Zhao, J. Tian, J. Fu, J. Chen, and J. Wen, “Feamix: Feature mix with memory batch based on self-consistency learning for code generation and code translation,” IEEE Transactions on Emerging Topics in Computational Intelligence , 2024

    2024

  18. [26]

    Exploring clean label backdoor attacks and defense in language models,

    S. Zhao, L. A. Tuan, J. Fu, J. Wen, and W. Luo, “Exploring clean label backdoor attacks and defense in language models,” IEEE/ACM Transactions on Audio, Speech, and Language Processing , 2024

    2024

  19. [27]

    Updating language models with unstructured facts: Towards practical knowledge editing,

    X. Wu, L. Pan, W. Y . Wang, and A. T. Luu, “Updating language models with unstructured facts: Towards practical knowledge editing,” 2024

    2024

  20. [28]

    Antileak-bench: Preventing data contamination by automatically constructing benchmarks with updated real-world knowledge,

    X. Wu, L. Pan, Y . Xie, R. Zhou, S. Zhao, Y . Ma, M. Du, R. Mao, A. T. Luu, and W. Y . Wang, “Antileak-bench: Preventing data contamination by automatically constructing benchmarks with updated real-world knowledge,” arXiv preprint arXiv:2412.13670 , 2024

    2024 arXiv

  21. [29]

    Are LLMs good zero-shot fallacy classifiers?

    F. Pan, X. Wu, Z. Li, and A. T. Luu, “Are LLMs good zero-shot fallacy classifiers?” arXiv preprint arXiv:2410.15050 , 2024

    2024 arXiv

  22. [30]

    Can textual gradient work in federated learning?

    M. Chen, R. Jin, W. Deng, Y . Chen, Z. Huang, H. Yu, and X. Li, “Can textual gradient work in federated learning?” in The Thirteenth International Conference on Learning Representations

  23. [31]

    A survey of large language models for healthcare: from data, technology, and applications to accountability and ethics,

    K. He, R. Mao, Q. Lin, Y . Ruan, X. Lan, M. Feng, and E. Cambria, “A survey of large language models for healthcare: from data, technology, and applications to accountability and ethics,” arXiv preprint arXiv:2310.05694, 2023

    arXiv

  24. [32]

    A survey on pragmatic processing techniques,

    R. Mao, M. Ge, S. Han, W. Li, K. He, L. Zhu, and E. Cambria, “A survey on pragmatic processing techniques,” Information Fusion , vol. 114, p. 102712, 2025

    2025

  25. [33]

    Few-shot fine-tuning sota summarization models for medical dialogues,

    D. F. Navarro, M. Dras, and S. Berkovsky, “Few-shot fine-tuning sota summarization models for medical dialogues,” in Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Student Research...

    2022

  26. [34]

    Parameter-efficient fine-tuning for large models: A comprehensive survey,

    Z. Han, C. Gao, J. Liu, J. Zhang, and S. Q. Zhang, “Parameter-efficient fine-tuning for large models: A comprehensive survey,” arXiv preprint arXiv:2403.14608, 2024

    2024 arXiv

  27. [35]

    A prompt pattern catalog to enhance prompt engineering with chatgpt,

    J. White, Q. Fu, S. Hays, M. Sandborn, C. Olea, H. Gilbert, A. Elnashar, J. Spencer-Smith, and D. C. Schmidt, “A prompt pattern catalog to enhance prompt engineering with chatgpt,” arXiv preprint arXiv:2302.11382, 2023

    2023 arXiv

  28. [36]

    Multi-dimensional evaluation of text summarization with in-context learning,

    S. Jain, V . Keshava, S. M. Sathyendra, P. Fernandes, P. Liu, G. Neubig, and C. Zhou, “Multi-dimensional evaluation of text summarization with in-context learning,” in The 61st Annual Meeting Of The Association For Computational Linguistics, 2023

    2023

  29. [37]

    From softmax to nucleusmax: A novel sparse language model for chinese radiology report summarization,

    S. Zhao, Q. Li, Y . Yang, J. Wen, and W. Luo, “From softmax to nucleusmax: A novel sparse language model for chinese radiology report summarization,” ACM Transactions on Asian and Low-Resource Language Information Processing , pp. 1–21, 2023

    2023

  30. [38]

    Usb: A unified summarization benchmark across tasks and domains,

    K. Krishna, P. Gupta, S. Ramprasad, B. C. Wallace, J. P. Bigham, and Z. C. Lipton, “Usb: A unified summarization benchmark across tasks and domains,” in Findings of EMNLP , 2023, pp. 8826–8845

    2023

  31. [39]

    Oasum: Large-scale open domain aspect-based summarization,

    X. Yang, K. Song, S. Cho, X. Wang, X. Pan, L. Petzold, and D. Yu, “Oasum: Large-scale open domain aspect-based summarization,” in Findings of ACL , 2023, pp. 4381–4401

    2023

  32. [40]

    Inducing document structure for aspect-based summarization,

    L. Frermann and A. Klementiev, “Inducing document structure for aspect-based summarization,” in Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, A. Korhonen, D. Traum, and L. Màrquez, Eds. Florence, Italy: Association for Computational L...

    2019

  33. [41]

    Meteor: An automatic metric for mt evaluation with improved correlation with human judgments,

    S. Banerjee and A. Lavie, “Meteor: An automatic metric for mt evaluation with improved correlation with human judgments,” in Proceedings of the acl workshop on intrinsic and extrinsic evaluation measures for machine translation and/or summarization , 2005, pp. 65–72

    2005

  34. [42]

    ROUGE: A package for automatic evaluation of summaries,

    C.-Y . Lin, “ROUGE: A package for automatic evaluation of summaries,” in Text Summarization Branches Out . Barcelona, Spain: Association for Computational Linguistics, Jul. 2004, pp. 74–81

    2004

  35. [43]

    Bertscore: Evaluating text generation with bert,

    T. Zhang, V . Kishore, F. Wu, K. Q. Weinberger, and Y . Artzi, “Bertscore: Evaluating text generation with bert,” arXiv preprint arXiv:1904.09675 , 2019

    1904 arXiv

  36. [44]

    The llama 3 herd of models,

    A. Dubey, A. Jauhri, A. Pandey, A. Kadian, A. Al-Dahle, A. Letman, A. Mathur, A. Schelten, A. Yang, A. Fan et al., “The llama 3 herd of models,” arXiv preprint arXiv:2407.21783 , 2024

    2024 arXiv

  37. [45]

    Mixtral of experts,

    A. Q. Jiang, A. Sablayrolles, A. Roux, A. Mensch, B. Savary, C. Bamford, D. S. Chaplot, D. d. l. Casas, E. B. Hanna, F. Bressand et al., “Mixtral of experts,” arXiv preprint arXiv:2401.04088 , 2024

    2024 arXiv

  38. [46]

    Compressing context to enhance inference efficiency of large language models,

    Y . Li, B. Dong, F. Guerin, and C. Lin, “Compressing context to enhance inference efficiency of large language models,” in Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing , H. Bouamor, J. Pino, and K. Bali, Eds. Singapore: Association for...

    2023

  39. [47]

    On context utilization in summarization with large language models,

    M. Ravaut, A. Sun, N. Chen, and S. Joty, “On context utilization in summarization with large language models,” in Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2024, pp. 2764–2781

    2024

  40. [48]

    Exploring the limits of chatgpt for query or aspect-based text summarization,

    X. Yang, Y . Li, X. Zhang, H. Chen, and W. Cheng, “Exploring the limits of chatgpt for query or aspect-based text summarization,” arXiv preprint arXiv:2302.08081, 2023

    2023 arXiv

  41. [49]

    Learning disentangled representation via domain adaptation for dialogue summarization,

    J. Li, Y . Xia, X. Cheng, D. Zhao, and R. Yan, “Learning disentangled representation via domain adaptation for dialogue summarization,” in Proceedings of the ACM Web Conference 2023 , 2023, pp. 1693–1702

    2023

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.