Pith. sign in

REVIEW 5 major objections 5 minor 1 cited by

Temporal reasoning for timeline summarisation in social media

T0 review · 5 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read The paper claims that temporal reasoning over event sequences, transferred from a large teacher to a small student by knowledge distillation, improves out-of-domain timeline summarisation and reduces hallucinations.

desk verdict A useful new dataset and a clean control, but the temporal-reasoning mechanism is unproven and the evaluation is too small to carry the claims. read the letter →

arxiv 2501.00152 v3 pith:NQXPT2EZ submitted 2024-12-30 cs.CL cs.AI

Jiayu Song , Mahmud Elahi Akhter , Dana Atzil Slonim , Maria Liakata This is my paper · ORCID
classification cs.CLcs.AI
keywords temporalreasoningtimelinesummarisationknowledgedistillationlargelanguagemodelsmentalhealthsocialmediaNarrativeReasonhallucinationreduction
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

The paper claims that temporal reasoning over full event sequences, not just event pairs, is a transferable skill that improves timeline summarisation in language models. It introduces NarrativeReason, a dataset of 19,614 temporal-relation question/answer pairs built from densely annotated narratives, fine-tunes a large teacher model on it, and distills the teacher into a smaller student while the student is simultaneously trained on news timeline summarisation. Applied to an out-of-domain mental-health timeline task, the distilled student produces summaries that score higher on factual consistency and evidence appropriateness than baselines including joint training, in-context learning, and a prior hierarchical VAE, and clinical raters prefer them. The paper also uses representation analyses to argue that distillation, unlike joint training, creates task-aligned internal representations that explain the improvement.

What carries the argument

The load-bearing mechanism is the two-stage distillation pipeline. NarrativeReason converts dense NarrativeTime annotations into event triples and full-sequence temporal-relation question/answer pairs, so the teacher learns relations among a series of events rather than isolated pairs. The teacher is fine-tuned with LoRA, then frozen; the student is trained on news timeline summarisation while three representation-matching losses, neuron selectivity transfer (NST), probabilistic knowledge transfer (PRT), and contrastive representation distillation (CRD), pull its hidden representations toward the teacher's. NST plus PRT combined gives the best results, and the authors argue these structure-matching losses are better suited to temporal coherence than CRD's instance discrimination.

What would settle it

Run the distilled student on an event-ordering benchmark, for example the event-event portion of TEMPREASON, and compare it with a student distilled from a teacher fine-tuned on a non-temporal task of the same format; equal performance would show the gains are not caused by temporal reasoning transfer.

Watch

Extended reading notes

Core claim

The central claim is that knowledge distillation is the right vehicle for combining temporal reasoning with timeline summarisation: fine-tuning on temporal reasoning alone, timeline summarisation alone, or both jointly all underperform, while a student that learns timeline summarisation and simultaneously matches a temporal-reasoning teacher's representations transfers the temporal skill without task interference. The empirical anchor is the out-of-domain TalkLife setting, where the best KD model with NST and PRT reaches a factual consistency of 0.438 and an evidence appropriateness of 0.973, and human raters give it the highest scores on factual consistency and general usefulness. The authors attribute the gain to the student learning more polysemantic, task-aligned representations, shown through UMAP and CKA analyses comparing the distilled student with the jointly trained model.

Load-bearing premise

The paper assumes the student's better summaries come from the event-ordering knowledge it absorbed from the teacher, but the student is never tested on event ordering directly, so other differences between teacher and student could explain the gains.

Editorial extensions

If this is right

  • A smaller model can acquire temporal reasoning from a larger teacher and apply it in a domain the teacher never saw, outperforming models fine-tuned on either task alone.
  • Joint training on temporal reasoning and timeline summarisation actively hurts performance, so distillation's separation of the two learning signals is necessary, not optional.
  • Matching structural activation and probability distributions (NST and PRT) beats contrastive representation matching (CRD) for temporally coherent summarisation.
  • The best distilled model's factual-consistency score of 0.438 substantially exceeds timeline-only distillation (0.330) and joint training (0.238), supporting the paper's claim that temporal reasoning specifically reduces hallucinations.

Reading between the lines

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

  • The causal story would be strengthened by a direct student-side test; the paper only verifies the teacher on TEMPREASON, so student-side temporal reasoning gains remain an inference.
  • The same recipe could transfer to other longitudinal summarisation settings with scarce in-domain labels, such as clinical notes or support conversations, where event order is clinically meaningful.
  • Because joint training fails while distillation succeeds, the teacher may be acting as a regulariser; a matched non-temporal reasoning teacher would reveal whether temporal content or general representation-matching drives the gain.
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

5 major / 5 minor

Summary. The paper proposes to improve timeline summarisation by first fine-tuning a large teacher LLM (LLaMA-3-8B) on a new temporal-reasoning dataset, NarrativeReason, built from NarrativeTime, and then distilling the teacher's knowledge into a smaller Phi-3 student while simultaneously fine-tuning the student on news timeline summarisation. The resulting student is applied out-of-domain to mental-health timeline summarisation on 30 TalkLife timelines, where it is compared with several baselines and ablations using Factual Consistency (FC), Evidence Appropriateness (EA), and a small human evaluation by two clinical psychology students. The central claim is that the KD approach 'produces more accurate summaries while significantly reducing hallucinations'.

Significance. If the central claim were fully supported, the paper would make a useful contribution: NarrativeReason is a new dataset with dense event-event temporal relations in narrative form, and the KD setup with timeline-summarisation-specific controls (KDtimeline, KDorigin) is a reasonable way to isolate the effect of temporal-reasoning fine-tuning. The out-of-domain evaluation on TalkLife is also a strength, as is the attempt to explain the gains through representation analyses (UMAP, CKA). However, the current evidence is not yet sufficient to establish the causal mechanism: the student is never evaluated on a temporal-reasoning task, the automatic results lack significance testing, and the evaluation protocol gives systems differential access to gold evidence. These issues affect the paper's main conclusion, so the manuscript needs major revision.

major comments (5)
  1. [§4.3, Table 2] The evaluation protocol is confounded by differential access to gold evidence. The paper states that 'we use the annotated timeline evidence from Song et al. (2024) to directly generate high-level summaries,' whereas the TH-VAE baseline must first extract evidence from timelines and then translate it to high-level summaries. The reported gains over TH-VAE (FC .378 to .438; EA .970 to .973) are therefore uninterpretable as evidence for temporal reasoning: the comparison mixes the effect of the proposed method with the effect of bypassing evidence extraction. A fair comparison would either provide the same gold evidence to all systems or evaluate end-to-end evidence extraction and summarisation for every system.
  2. [§5, Table 2] All automatic conclusions rest on single-point FC and EA values computed over only 30 TalkLife timelines, with no confidence intervals, bootstrap estimates, or paired significance tests. Differences such as FC .424 vs .438 and EA .971 vs .973 are plausibly within noise. The abstract and conclusion claim that the KD approach 'significantly reduces hallucinations,' but 'significantly' is not supported by any statistical test. Per-timeline paired tests or bootstrap confidence intervals should be reported before making the headline claim.
  3. [§3.3, Appendix A.2] The paper's causal mechanism is that temporal-reasoning knowledge is distilled into the student, yet the student is never evaluated on a temporal-reasoning benchmark. Appendix A.2 reports that fine-tuning improves the teacher on TEMPREASON L3 by +0.07 F1, but no equivalent evaluation is given for P-Phi or L-Phi. The observed gains on TalkLife could therefore arise from other teacher properties that are correlated with NarrativeReason SFT, such as better instruction following or different output style, rather than from temporal reasoning per se. A direct student-side temporal-reasoning evaluation is load-bearing for the central claim and should be added.
  4. [§5, Table 1] The best KD configuration is selected using the same 30 TalkLife timelines that are later used for the final evaluation. Specifically, Table 1 compares several KD strategies on FC and EA, and NST&PRT is chosen as the best; the same 30 timelines then appear in Table 2 as the headline result. This is test-set selection and can inflate the reported numbers. A held-out validation split, nested evaluation, or at least an explicit acknowledgment of this selection effect is needed.
  5. [§5, Table 3] The human evaluation uses only two annotators, each rating 30 summaries on 5-point Likert scales, with no inter-annotator reliability measure and no significance testing. The paper states that clinical psychologists have 'a clear preference' for L-Phi, but the Likert differences are small (e.g., MOC 3.47 vs 3.42 for LLaMA) and the claim is not statistically supported. Reporting Cohen's kappa or similar, and at least paired tests on per-item ratings, would be necessary to support the preference claim.
minor comments (5)
  1. [§5, Table 1] The acronym is used inconsistently: the methodology defines PRT, but the results section says 'PKT performed the best.' Please unify the notation throughout.
  2. [Table 2] The row label 'KDtimline' is a typo; it should read 'KDtimeline.' Also, the column header 'Data for Fine-Tuning Teacher Model' is redundant and could be simplified.
  3. [Eq. (3)] There is a typo in 'negtive' in the CRD loss equation, and the text says h should be normalized by L-2 norm but the displayed formula does not show the normalization explicitly.
  4. [Appendix A.4] The example outputs contain apparent generation artifacts such as 'mental wellbes' and 'life-thyesing thoughts.' If these are verbatim model outputs, they should be clearly marked as such; otherwise the appendix should be proofread.
  5. [§4.1] The paper does not report how the 5,000 Wikipedia timelines are split into training and validation sets, nor how teacher and student training epochs and checkpoints were selected. Please add these details.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the distillation pipeline is tested against external controls and an out-of-domain benchmark, with no load-bearing step reducing to its inputs.

full rationale

The paper's central claim is that temporal reasoning acquired on NarrativeReason improves out-of-domain timeline summarisation after knowledge distillation. The derivation chain is empirically controlled rather than definitional. The teacher's temporal reasoning improvement is measured on TEMPREASON, an external benchmark, not on the training distribution itself (Appendix A.2). The contribution of the NarrativeReason-tuned teacher is isolated by the KDtimeline and KDorigin ablations in Table 2, which differ only in what the teacher was fine-tuned on; the student and evaluation data are held constant. The TalkLife evaluation is out-of-domain with respect to both teacher and student training, and includes human evaluation by clinical psychology graduate students. The self-citations to Song et al. (2024) and Tsakalidis et al. (2022) provide the task format, gold summaries, evidence annotations, and the TH-VAE baseline; these are published data resources and baselines, not a cited theorem that forces the result. A genuine weakness is that the student is never directly evaluated on a temporal reasoning task, so the causal interpretation that distilled temporal reasoning drives the gains is not fully established; however, this is an evidential gap, not a circular reduction. No equation or fitted parameter in the paper is equivalent by construction to the claimed prediction, and no load-bearing argument reduces to a self-citation chain. Therefore the paper is self-contained against external benchmarks and receives a circularity score of 0.

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

The paper does not introduce any fundamentally new physical or conceptual entities such as a new force or particle. It does rely on domain assumptions about the sufficiency of a closed set of temporal relations, the transferability of teacher knowledge through distillation, the validity of cross-domain training, and the use of gold evidence as a proxy for the full summarisation task.

assumptions (4)
  • domain assumption Temporal relations (BEFORE, AFTER, INCLUDES, IS_INCLUDED, SIMULTANEOUS) are sufficient to represent narrative timelines.
    The NarrativeReason dataset is constructed using these relation types from NarrativeTime annotations; if some event-order information is not captured by this closed set, the dataset will be incomplete.
  • domain assumption Fine-tuning a large LM on NarrativeReason improves its temporal reasoning, and this improvement transfers via KD to a smaller model.
    The teacher's measured improvement on TEMPREASON L3 is only 0.07 F1, and the student is never evaluated on a temporal reasoning benchmark, so the transfer step is assumed rather than demonstrated.
  • domain assumption Training on Wikipedia-derived news timeline summaries is a valid source for out-of-domain transfer to mental-health social media timelines.
    Used in §4.1 to justify the training data choice; the paper does not test whether other training domains would work better.
  • ad hoc to paper The gold evidence annotations from Song et al. (2024) can be used directly to generate high-level summaries.
    In §4.3 the authors state they directly use the annotated evidence instead of extracting it, which changes the task relative to the TH-VAE baseline and is an assumption made for this paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Temporal reasoning for timeline summarisation in social media." pith.science (2026). https://pith.science/paper/NQXPT2EZ

@misc{pith2026250100152,
  author       = {Pith},
  title        = {Pith review of: Temporal reasoning for timeline summarisation in social media},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NQXPT2EZ}},
  note         = {Machine review of arXiv:2501.00152}
}
read the original abstract

This paper explores whether enhancing temporal reasoning capabilities in Large Language Models (LLMs) can improve the quality of timeline summarisation, the task of summarising long texts containing sequences of events, such as social media threads. We first introduce NarrativeReason, a novel dataset focused on temporal relationships among sequential events within narratives, distinguishing it from existing temporal reasoning datasets that primarily address pair-wise event relationships. Our approach then combines temporal reasoning with timeline summarisation through a knowledge distillation framework, where we first fine-tune a teacher model on temporal reasoning tasks and then distill this knowledge into a student model while simultaneously training it for the task of timeline summarisation. Experimental results demonstrate that our model achieves superior performance on out-of-domain mental health-related timeline summarisation tasks, which involve long social media threads with repetitions of events and a mix of emotions, highlighting the importance and generalisability of leveraging temporal reasoning to improve timeline summarisation.

Figures

Figures reproduced from arXiv: 2501.00152 by the authors.

Figure 1
Figure 1. Overview of proposed method. (a) represents fine-tuning the teacher model on the temporal reasoning [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The temporal relations between events. The text in the left column comes from the [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. The UMAP projection forPhijoint and L-Phi show the last layer activations for both models. We can see that L-Phi has more polysemantic activations compared to Phijoint [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: CKA similarity score of both within and between L-Phi and Phi [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. TIM: A Large-Scale Dataset and large Timeline Intelligence Model for Open-domain Timeline Summarization

    cs.CL 2025-06 conditional novelty 6.0 of 10

    A timeline summarization model trained on a new 1,189-topic dataset outperforms much larger general LLMs on the authors' open-domain test set.

Reference graph

Works this paper leans on

50 extracted references · 20 canonical work pages · cited by 1 Pith paper

  1. [1]

    URL: " 'urlintro :=

    ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Bertram C Bruce. 1972. A model for temporal references and its application in a question answering program. Artificial intelligence, 3:1--25

    work page 1972

  4. [4]

    Taylor Cassidy, Bill McDowell, Nathanael Chambers, and Steven Bethard. 2014. https://doi.org/10.3115/V1/P14-2082 An annotation framework for dense event ordering . In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics, ACL 2014, June 22-27, 2014, Baltimore, MD, USA, Volume 2: Short Papers , pages 501--506. The Associat...

    doi:10.3115/v1/p14-2082 2014

  5. [5]

    Chunkit Chan, Cheng Jiayang, Weiqi Wang, Yuxin Jiang, Tianqing Fang, Xin Liu, and Yangqiu Song. 2024. Exploring the potential of chatgpt on sentence level relations: A focus on temporal, causal, and discourse relations. In Findings of the Association for Computational Linguistics: EACL 2024, pages 684--721

    work page 2024

  6. [6]

    Meiqi Chen, Yubo Ma, Kaitao Song, Yixin Cao, Yan Zhang, and Dongsheng Li. 2024. https://doi.org/10.18653/V1/2024.ACL-LONG.512 Improving large language models in event relation logical prediction . In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), ACL 2024, Bangkok, Thailand, August 11-16, 2...

    doi:10.18653/v1/2024.acl-long.512 2024

  7. [7]

    Xiuying Chen, Mingzhe Li, Shen Gao, Zhangming Chan, Dongyan Zhao, Xin Gao, Xiangliang Zhang, and Rui Yan. 2023. Follow the timeline! generating an abstractive and extractive timeline summary in chronological order. ACM Transactions on Information Systems, 41(1):1--30

    work page 2023

  8. [8]

    Yao Cheng, Peter Anick, Pengyu Hong, and Nianwen Xue. 2013. Temporal relation discovery between events and temporal expressions identified in clinical narrative. Journal of biomedical informatics, 46:S48--S53

    work page 2013

Show all 50 references
  1. [9]

    Zheng Chu, Jingchang Chen, Qianglong Chen, Weijiang Yu, Haotian Wang, Ming Liu, and Bing Qin. 2024. https://doi.org/10.18653/V1/2024.ACL-LONG.66 Timebench: A comprehensive evaluation of temporal reasoning abilities in large language models . In Proceedings of the 62nd Annual M...

    2024 doi

  2. [10]

    Alexis Conneau, Shijie Wu, Haoran Li, Luke Zettlemoyer, and Veselin Stoyanov. 2020. https://doi.org/10.18653/v1/2020.acl-main.536 Emerging cross-lingual structure in pretrained language models . In Proceedings of the 58th Annual Meeting of the Association for Computational Lin...

    2020 doi

  3. [11]

    Maksym Del and Mark Fishel. 2021. http://arxiv.org/abs/2109.01207 Establishing interlingua in multilingual language models . CoRR, abs/2109.01207

    2021 arXiv

  4. [12]

    Yu Feng, Ben Zhou, Haoyu Wang, Helen Jin, and Dan Roth. 2023. https://doi.org/10.18653/V1/2023.ACL-LONG.671 Generic temporal reasoning with differential analysis and explanation . In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volum...

    2023 doi

  5. [13]

    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. 2022. https://openreview.net/forum?id=nZeVKeeFYf9 Lora: Low-rank adaptation of large language models . In The Tenth International Conference on Learning Representat...

    2022

  6. [14]

    Qisheng Hu, Geonsik Moon, and Hwee Tou Ng. 2024. https://doi.org/10.18653/v1/2024.acl-long.390 From moments to milestones: Incremental timeline summarization leveraging large language models . In Proceedings of the 62nd Annual Meeting of the Association for Computational Lingu...

    2024 doi

  7. [15]

    Rikui Huang, Wei Wei, Xiaoye Qu, Shengzhe Zhang, Dangyang Chen, and Yu Cheng. 2024. https://aclanthology.org/2024.acl-long.580 Confidence is not timeless: Modeling temporal validity for rule-based temporal knowledge graph forecasting . In Proceedings of the 62nd Annual Meeting...

    2024

  8. [16]

    Zehao Huang and Naiyan Wang. 2017. Like what you like: Knowledge distill via neuron selectivity transfer. arXiv preprint arXiv:1707.01219

    2017 arXiv

  9. [17]

    Raghav Jain, Daivik Sojitra, Arkadeep Acharya, Sriparna Saha, Adam Jatowt, and Sandipan Dandapat. 2023. https://doi.org/10.18653/v1/2023.emnlp-main.418 Do language models have a common sense regarding time? revisiting temporal commonsense reasoning in the era of large language...

    2023 doi

  10. [18]

    Hyuckchul Jung, James Allen, Nate Blaylock, William de Beaumont, Lucian Galescu, and Mary Swift. 2011 a . https://aclanthology.org/W11-0219 Building timelines from narrative clinical records: Initial results based-on deep natural language understanding . In Proceedings of B io...

    2011

  11. [19]

    Hyuckchul Jung, James Allen, Nate Blaylock, William de Beaumont, Lucian Galescu, and Mary Swift. 2011 b . Building timelines from narrative clinical records: initial results based-on deep natural language understanding. In Proceedings of BioNLP 2011 workshop, pages 146--154

    2011

  12. [20]

    Daniel Khashabi. 2019. Reasoning-Driven Question-Answering for Natural Language Understanding. University of Pennsylvania

    2019

  13. [21]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. 2015. http://arxiv.org/abs/1412.6980 Adam: A method for stochastic optimization . In 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings

    2015 arXiv

  14. [22]

    Simon Kornblith, Mohammad Norouzi, Honglak Lee, and Geoffrey E. Hinton. 2019. http://proceedings.mlr.press/v97/kornblith19a.html Similarity of neural network representations revisited . In Proceedings of the 36th International Conference on Machine Learning, ICML 2019, 9-15 Ju...

    2019

  15. [23]

    Manling Li, Tengfei Ma, Mo Yu, Lingfei Wu, Tian Gao, Heng Ji, and Kathleen R. McKeown. 2021. https://doi.org/10.18653/V1/2021.EMNLP-MAIN.519 Timeline summarization based on event graph compression via time-aware optimal transport . In Proceedings of the 2021 Conference on Empi...

    2021 doi

  16. [24]

    Leland McInnes, John Healy, Nathaniel Saul, and Lukas Großberger. 2018. https://doi.org/10.21105/joss.00861 Umap: Uniform manifold approximation and projection . Journal of Open Source Software, 3(29):861

    2018 doi

  17. [25]

    Ibraheem Muhammad Moosa, Mahmud Elahi Akhter, and Ashfia Binte Habib. 2023. https://doi.org/10.18653/v1/2023.findings-eacl.50 Does transliteration help multilingual language modeling? In Findings of the Association for Computational Linguistics: EACL 2023, pages 670--685, Dubr...

    2023 doi

  18. [26]

    Benjamin Muller, Yanai Elazar, Beno \^ t Sagot, and Djam \'e Seddah. 2021. https://doi.org/10.18653/v1/2021.eacl-main.189 First align, then predict: Understanding the cross-lingual ability of multilingual BERT . In Proceedings of the 16th Conference of the European Chapter of ...

    2021 doi

  19. [27]

    Alexander Nakhimovsky. 1987. https://aclanthology.org/E87-1042 Temporal reasoning in natural language understanding: The temporal structure of the narrative . In Third Conference of the E uropean Chapter of the Association for Computational Linguistics , Copenhagen, Denmark. A...

    1987

  20. [28]

    Qiang Ning, Hao Wu, Rujun Han, Nanyun Peng, Matt Gardner, and Dan Roth. 2020. https://doi.org/10.18653/v1/2020.emnlp-main.88 TORQUE : A reading comprehension dataset of temporal ordering questions . In Proceedings of the 2020 Conference on Empirical Methods in Natural Language...

    2020 doi

  21. [29]

    Qiang Ning, Ben Zhou, Zhili Feng, Haoruo Peng, and Dan Roth. 2018. https://doi.org/10.18653/v1/D18-2013 C og C omp T ime: A tool for understanding time in natural language . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing: System Demon...

    2018 doi

  22. [30]

    Chris Olah, Nick Cammarata, Ludwig Schubert, Gabriel Goh, Michael Petrov, and Shan Carter. 2020. https://doi.org/10.23915/distill.00024.001 Zoom in: An introduction to circuits . Distill. Https://distill.pub/2020/circuits/zoom-in

    2020 doi

  23. [31]

    Nikolaos Passalis and Anastasios Tefas. 2018. Learning deep representations with probabilistic knowledge transfer. In Proceedings of the European Conference on Computer Vision (ECCV), pages 268--284

    2018

  24. [32]

    James Pustejovsky, Patrick Hanks, Roser Sauri, Andrew See, Robert Gaizauskas, Andrea Setzer, Dragomir Radev, Beth Sundheim, David Day, Lisa Ferro, et al. 2003. The timebank corpus. In Corpus linguistics, volume 2003, page 40. Lancaster, UK

    2003

  25. [33]

    Yifu Qiu, Zheng Zhao, Yftah Ziser, Anna Korhonen, Edoardo M Ponti, and Shay B Cohen. 2023. Are large language models temporally grounded? arXiv preprint arXiv:2311.08398

    2023 arXiv

  26. [34]

    Hossein Rajaby Faghihi, Bashar Alhafni, Ke Zhang, Shihao Ran, Joel Tetreault, and Alejandro Jaimes. 2022. https://doi.org/10.18653/v1/2022.findings-emnlp.400 C risis LTLS um: A benchmark for local crisis event timeline extraction and summarization . In Findings of the Associat...

    2022 doi

  27. [35]

    Anna Rogers, Marzena Karpinska, Ankita Gupta, Vladislav Lialin, Gregory Smelkov, and Anna Rumshisky. 2024. https://aclanthology.org/2024.lrec-main.1054 Narrativetime: Dense temporal annotation on a timeline . In Proceedings of the 2024 Joint International Conference on Computa...

    2024

  28. [36]

    Tim Sainburg, Leland McInnes, and Timothy Q. Gentner. 2021. https://doi.org/10.1162/NECO\_A\_01434 Parametric UMAP embeddings for representation and semisupervised learning . Neural Comput., 33(11):2881--2907

    2021 doi

  29. [37]

    Jiayu Song, Jenny Chim, Adam Tsakalidis, Julia Ive, Dana Atzil - Slonim, and Maria Liakata. 2024. https://doi.org/10.18653/V1/2024.FINDINGS-ACL.873 Combining hierachical vaes with llms for clinically meaningful timeline summarisation in social media . In Findings of the Associ...

    2024 doi

  30. [38]

    Julius Steen and Katja Markert. 2019. https://doi.org/10.18653/v1/D19-5403 Abstractive timeline summarization . In Proceedings of the 2nd Workshop on New Frontiers in Summarization, pages 21--31, Hong Kong, China. Association for Computational Linguistics

    2019 doi

  31. [39]

    Qingyu Tan, Hwee Tou Ng, and Lidong Bing. 2023. https://doi.org/10.18653/V1/2023.ACL-LONG.828 Towards benchmarking and improving the temporal reasoning capability of large language models . In Proceedings of the 61st Annual Meeting of the Association for Computational Linguist...

    2023 doi

  32. [40]

    Zineng Tang, Jaemin Cho, Hao Tan, and Mohit Bansal. 2021. https://proceedings.neurips.cc/paper/2021/hash/ccdf3864e2fa9089f9eca4fc7a48ea0a-Abstract.html Vidlankd: Improving language understanding via video-distilled knowledge transfer . In Advances in Neural Information Process...

    2021

  33. [41]

    Yonglong Tian, Dilip Krishnan, and Phillip Isola. 2019. Contrastive representation distillation. arXiv preprint arXiv:1910.10699

    2019 arXiv

  34. [42]

    Adam Tsakalidis, Federico Nanni, Anthony Hills, Jenny Chim, Jiayu Song, and Maria Liakata. 2022. https://doi.org/10.18653/v1/2022.acl-long.318 Identifying moments of change from longitudinal user text . In Proceedings of the 60th Annual Meeting of the Association for Computati...

    2022 doi

  35. [43]

    Talia Tseriotou, Adam Tsakalidis, Peter Foster, Terence Lyons, and Maria Liakata. 2023. https://doi.org/10.18653/v1/2023.findings-acl.310 Sequential path signature networks for personalised longitudinal language modeling . In Findings of the Association for Computational Lingu...

    2023 doi

  36. [44]

    Siddharth Vashishtha, Adam Poliak, Yash Kumar Lal, Benjamin Van Durme, and Aaron Steven White. 2020. https://doi.org/10.18653/v1/2020.findings-emnlp.363 Temporal reasoning in natural language inference . In Findings of the Association for Computational Linguistics: EMNLP 2020,...

    2020 doi

  37. [45]

    Yuqing Wang and Yun Zhao. 2024. https://doi.org/10.18653/V1/2024.FINDINGS-ACL.382 TRAM: benchmarking temporal reasoning for large language models . In Findings of the Association for Computational Linguistics, ACL 2024, Bangkok, Thailand and virtual meeting, August 11-16, 2024...

    2024 doi

  38. [47]

    Georg Wenzel and Adam Jatowt. 2023 b . https://doi.org/10.48550/ARXIV.2308.00002 An overview of temporal commonsense reasoning and acquisition . CoRR, abs/2308.00002

    2023 doi arXiv

  39. [48]

    Bowen Xing and Ivor W. Tsang. 2023. https://doi.org/10.1109/TPAMI.2023.3289509 Relational temporal graph reasoning for dual-task dialogue language understanding . IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(11):13170--13184

    2023

  40. [49]

    Siheng Xiong, Ali Payani, Ramana Kompella, and Faramarz Fekri. 2024. https://doi.org/10.18653/V1/2024.ACL-LONG.563 Large language models can learn temporal reasoning . In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Pa...

    2024 doi

  41. [50]

    Xinliang Frederick Zhang, Nick Beauchamp, and Lu Wang. 2024. Narrative-of-thought: Improving temporal reasoning of large language models via recounted narratives. arXiv preprint arXiv:2410.05558

    2024 arXiv

  42. [51]

    going on a vacation

    Ben Zhou, Daniel Khashabi, Qiang Ning, and Dan Roth. 2019. https://doi.org/10.18653/V1/D19-1332 "going on a vacation" takes longer than "going for a walk": A study of temporal commonsense understanding . In Proceedings of the 2019 Conference on Empirical Methods in Natural Lan...

    2019 doi

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.