Pith. sign in

REVIEW 3 major objections 5 minor 43 references

Abstract Event Causal Rules: Induction and Application

T0 review · 3 major / 5 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read This paper claims that abstracting concrete cause-effect event pairs into generalized causal rules, retrieved and injected into a graph-based event predictor, improves event prediction accuracy on rare and unseen events.

desk verdict Novel relation-level causal abstraction idea, but the main generalization claim is compromised by data leakage: rule induction and retriever training use the full datasets, including the downstream test events. read the letter →

arxiv 2608.05205 v1 pith:GMPK46BJ submitted 2026-08-05 cs.AI

Ziwei Zheng , Peiqiong Chen , Bang Wang This is my paper · ORCID
classification cs.AI
keywords causalknowledgeabstractioneventrulespredictioncausalitygraphmulti-agentLLMbaseinductionrule-guidedattentionlong-tailgeneralization
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 sets out to show that event causality can be represented at the level of abstract rules, rather than concrete instance-level pairs, and that such rules transfer across lexically distinct events. It proposes Abstract Event Causal Rules (AECR), relation-level abstractions like 'natural disaster to property damage' distilled from many concrete cause-effect pairs by a five-agent LLM pipeline with similarity-constrained clustering. To demonstrate utility, it builds two AECR knowledge bases and injects retrieved rules into the Causality Graph Event Prediction task via a rule-guided attention encoder with gated fusion. The central empirical claim is that applying these rules yields consistent gains over state-of-the-art baselines, with the largest improvements on rare and unseen events. If true, this gives event-centric systems a reusable causal prior that works where instance-level knowledge fails.

What carries the argument

The load-bearing mechanism is paired causal abstraction: the cause and effect are processed as one unit so the causal linkage survives abstraction, and the CACI multi-agent system (restatement, gatekeeping, abstraction, selection, judgment) produces validated concrete causal logics via an iterative feedback loop. These logics are embedded with a frozen RoBERTa encoder and grouped by agglomerative hierarchical clustering under an intra-cluster similarity threshold, and each cluster is distilled by an LLM into a single Abstract Event Causal Rule (AECR), a transferable causal pattern linking parent-concepts such as 'natural disaster' and 'property damage'. On the application side, the AR-GCAE encoder encodes the event graph with a topology-aware transformer, uses the [MASK] representation of the unknown effect to retrieve the top-K rules through a trained MLM-style retriever, re-encodes the graph with the rule texts attached, and fuses rule-free and rule-guided representations with a learned gate.

What would settle it

Re-run the CACI induction on the training split alone, keeping the downstream test pairs out of the clustering and rule-distillation steps, then re-evaluate AR-GCAE; if the rare and unseen gains in Fig. 5 shrink or disappear, the reported generalization is a rule-level memory effect.

Watch

Extended reading notes

Core claim

The paper's central discovery is that causal abstraction should take the complete causal relation, the paired cause and effect, as the unit of abstraction, not each event in isolation. Its multi-agent CACI pipeline jointly restates, filters, abstracts, selects, and judges abstracted cause-effect logics until the causal force survives, then clusters these logics and distills each cluster into one Abstract Event Causal Rule. The paper reports that injecting the retrieved rules into a topology-aware transformer through gated fusion yields consistent gains over strong baselines on two event-prediction benchmarks, with the largest improvements on rare and unseen events, and that the rules transfer across datasets.

Load-bearing premise

The generalization claim rests on the assumption that the abstract rules were induced from the training pairs only, so gains on never-seen test events reflect genuine transfer rather than the rules having memorized those specific events.

Editorial extensions

If this is right

  • Event prediction systems gain a usable causal prior: rare and wholly unseen consequence events become predictable from abstract logic rather than lexical co-occurrence.
  • Abstract rules transfer across corpora and domains: rules built from MAVEN improve ESC prediction and vice versa, so causal knowledge is reusable rather than dataset-specific.
  • Less task-specific training data is needed: with 60-80% of the training data, AR-GCAE already matches or beats fully trained strong baselines.
  • The injected knowledge is cheap and safe: rule guidance adds negligible compute, and random or wrong rules do not degrade accuracy thanks to the gated fusion.

Reading between the lines

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

  • Beyond the paper: the paired-abstraction principle should transfer to other relational semantics, such as temporal ordering or enablement, where abstracting one argument alone would equally sever the relation.
  • Beyond the paper: retrieved AECRs double as compact, human-readable rationales for a predicted consequence, offering a built-in explanation layer for risk-warning applications.
  • Beyond the paper: a cheaper variant worth testing is training a single student model on the accepted and rejected traces of the CACI multi-agent loop, potentially preserving rule quality while cutting the inference cost of five cooperating agents.
  • Beyond the paper: the weak effect of random rules suggests the framework would tolerate noisy or partially incorrect rule bases, so an experiment injecting rules from an unrelated domain is a natural next step.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes Abstract Event Causal Rules (AECRs), a relation-level causal abstraction in which concrete cause-effect event pairs are transformed into generalized causal rules by a multi-agent LLM pipeline (CACI), clustered under similarity constraints, and organized into knowledge bases. A retriever is trained to select relevant rules for a concrete event pair, and the retrieved rules are injected into a graph-based event prediction encoder (AR-GCAE) for the CGEP benchmark. The paper reports human evaluation of rule quality, retriever performance, main CGEP results, frequency-stratified results, cross-dataset transfer, low-resource curves, and ablations, and claims that AECRs deliver consistent gains, especially on rare and unseen events.

Significance. If the empirical claims held, the paper would make a useful contribution: the AECR formalism is clearly described, the multi-agent induction pipeline is a plausible way to obtain abstract causal knowledge, and the idea of evaluating abstract rules through a downstream event-prediction task is sensible. The human evaluation of rule reasonableness, discriminability, and usability is a real strength, as is the explicit cross-dataset transfer experiment. The central empirical claim, however, is not currently supported because the rule-induction and retriever-training procedures appear to use the full datasets, including the downstream test and 'Unseen' events. Since the headline result is that AECRs improve generalization to rare and unseen events, the leakage concern is load-bearing and must be resolved before the paper's main conclusion can be accepted.

major comments (3)
  1. [Section 4.1, Table 2; Section 6.1] The rule-induction procedure appears to use all causal pairs in the full datasets, while the downstream CGEP evaluation later splits the same datasets into train, development, and test sets. Section 4.1 states that CACI processes 32,263 (MAVEN) and 6,875 (ESC) 'original causal event pairs extracted from the ECGs' and reports no exclusion of downstream test or development pairs, and Section 6.1 describes the CGEP train/dev/test split without stating that rule induction was restricted to the training split. As written, a test event used in Fig. 5 can have contributed its abstracted causal logic to a cluster and thus to an abstract rule in the knowledge base. The 'Unseen (0)' performance gain in Fig. 5 would then be a rule-level memory effect rather than evidence of generalization.
  2. [Section 4.3, Section 3.3] The AECR retriever is trained on concrete event pairs together with the rules assigned by the same clustering that generated the rules, and these pairs are split 8:1:1 internally rather than according to the CGEP split. Consequently, a downstream test event can appear as a training sample for the frozen retriever that is later used to select rules for that same test event during AR-GCAE inference. This compounds the Section 4.1 leakage: the retriever is not blind to test instances, so the rule-selection step in Fig. 5 is contaminated as well. The retriever comparison in Table 5 is also somewhat circular because the training labels are the clustering's own assignments, but this latter point is secondary to the downstream contamination.
  3. [Section 6.3, Fig. 5; Section 6.5, Fig. 6] The frequency-stratified and low-resource results are the main evidence for the paper's central claim, and both depend directly on the leakage described above. 'Unseen (0)' events are defined as absent from the CGEP training split, but they are not necessarily absent from rule induction or from retriever training; therefore the statement that AR-GCAE 'retains a prominent, consistent performance advantage' on the Unseen subset cannot be attributed to abstract causal generalization. The cross-dataset experiment in Table 8 is less affected for the target dataset because the target test events are not in the source rule base, but it does not repair the primary experiments, and the paper does not report a re-run of the main evaluation with a clean split.
minor comments (5)
  1. [General] The dataset name is typeset inconsistently as 'MA VEN' in Table 2 and in several places in the text; it should be 'MAVEN' throughout.
  2. [Section 3.3] The similarity-based loss L_mod introduces a hyperparameter alpha, but the paper never reports the value of alpha, the temperature tau, or the L2 regularization coefficient lambda; these should be listed in Section 6.1 or in an appendix.
  3. [Section 5.2] The retrieval prompt uses '[PH]' as a placeholder for the unknown effect event and '[MASK]' for the rule position; it would aid reproducibility to specify whether these tokens are the model's existing special tokens or newly added vocabulary items.
  4. [Section 4.3] The sentence 'It is not unexpected that our retriever outperforms the other two, since it is specially trained on our constructed AECR knowledge base' correctly acknowledges the comparison is not apples-to-apples, but it also makes the retriever evaluation less informative; a clearer statement of what the SBERT and LLM baselines are meant to show would help.
  5. [Figures 2 and 4] Figures 2 and 4 are very dense, and the distinction between the rule-registration path, the retrieval path, and the training-objective path is hard to follow; adding a small worked example near the figures would improve readability.

Circularity Check

2 steps flagged · score 6.0 of 10

Rule induction and retriever training appear to use all CGEP pairs including downstream test events, so the 'Unseen' gains in Fig. 5 partially reduce to rule-level memorization rather than generalization.

  1. fitted input called prediction [Section 4.1 (Table 2) with Section 6.1 and Section 6.3]
    "Applying our CACI framework described in Section 3 to the two datasets, we obtain two AECR knowledge bases ... For MAVEN-CGEP, the CACI processes 32,263 original causal event pairs extracted from the ECGs. ... The remaining 26,084 pairs undergo abstraction and clustering, yielding 733 abstract event causal rules. ... Unseen (0, meaning the event never appears in training data). ... We adopt the same partitioning protocol as prior work [35] to ensure a fair comparison."

    Rule induction is described as consuming all original causal event pairs of each dataset, with no statement that the CGEP test, development, or held-out instances were excluded from the CACI abstraction and clustering process. The 'Unseen (0)' group in Fig. 5 is defined only by absence from CGEP training data, not by absence from rule induction. A test event can therefore be one of the 32,263 pairs whose causal logic was clustered and distilled into the AECR knowledge base. The reported gain on Unseen events is then partly a rule-memorization effect: the abstract rule for that test event is available by construction from the very pair being predicted, so the experiment does not cleanly measure transfer to unseen causal content.

  2. fitted input called prediction [Section 3.3 and Section 4.3]
    "To train this retriever, we construct training samples from the rule generation process described in Section 3.2: a causal event pair p_i=(e^c_i,e^e_i) is assigned an abstract rule r_j if it falls into a valid rule cluster, and these qualified pairs (p_i,r_i) serve as the data source for retriever training. ... To train and evaluate the retriever, we take the concrete event pairs together with their corresponding rules in each knowledge base and split them into training, validation, and test sets at an 8:1:1 ratio."

    The retriever's supervision labels are the cluster/rule assignments produced from the full set of concrete pairs in each knowledge base, and the 8:1:1 split is internal to that rule-assignment dataset. This internal split is not aligned with the CGEP train/dev/test partition used in Section 6.1. Consequently, a CGEP test event can have its own causal pair included in the retriever's training portion. When AR-GCAE later 'retrieves' rules for that event, the rule has already been learned as that event's own training label. The 'transferable causal prior' claimed for unseen events is therefore partly self-referential: the retrieval is reproducing rule assignments learned from the very test pairs being predicted.

full rationale

The central empirical claim—that AECRs improve generalization, especially on rare and unseen events—is partially circular because the rule base and the retriever are built from the full CGEP datasets without a stated exclusion of the downstream test partition. Section 4.1 reports applying CACI to all 32,263 (MAVEN) and 6,875 (ESC) original causal pairs; Section 6.1 introduces the CGEP train/dev/test split without saying that the rule-induction process withheld test events; and Section 4.3 trains the retriever on the concrete pairs 'together with their corresponding rules' using an internal 8:1:1 split. Together these passages show that a test event marked 'Unseen (0)' in Fig. 5 may have contributed its abstracted causal logic to a cluster/rule and may also have been a retriever training sample. The frequency-stratified and low-resource results are thus contaminated by rule-level memorization, so the 'predictions' on unseen events reduce in part to fitted inputs. I did not find load-bearing self-citation circularity: the same-author baselines [35]-[37] are used as competitors rather than as justification for the AECR framework, and the cross-dataset transfer experiment (Table 8) provides some independent evidence by applying rules from one dataset to a different target dataset. However, that cleaner evidence does not rescue the in-dataset 'Unseen' claims that the abstract and Fig. 5 emphasize. Overall, the circularity is partial rather than total: the rule construction itself is a real abstraction process, and the cross-dataset and human-evaluation results are not reducible to the leak, so a score of 6 is appropriate.

Assumptions & free parameters 5 free parameters · 3 assumptions · 0 invented entities

The central claim rests on several tuned hyperparameters and on the assumption that LLM-driven abstraction and embedding-based clustering preserve causal content. The most consequential dependence is the use of the full datasets for rule induction, which is a methodological choice rather than a tuneable parameter, and it creates the data-leakage risk that undermines the generalization evidence.

free parameters (5)
  • Intra-cluster similarity threshold tau = not reported
    Controls cluster cohesion in agglomerative hierarchical clustering (Section 3.2); no value or tuning procedure given in the main text.
  • Minimum cluster size mu = not reported
    Clusters with fewer than mu members are discarded as weakly generalizable (Section 3.2); the value affects how many abstract rules are produced (733 and 180 rules in Table 2).
  • Top-K retrieved rules = 3
    Number of abstract rules injected into AR-GCAE (Section 6.1); shown in Fig. 7 to be optimal around K=3.
  • Retriever loss hyperparameters lambda, temperature tau, alpha = not reported
    Hyperparameters in the retriever training loss in Section 3.3; no values are provided.
  • Learning rates and epochs = MAVEN: lr=1e-6, 10 epochs; ESC: lr=5e-6, 15 epochs
    Reported in Section 6.1 as tuned on the development set.
assumptions (3)
  • domain assumption LLM agents reliably abstract concrete events into parent-concepts while preserving causal validity
    The CACI pipeline (Section 3.1) delegates restatement, gating, abstraction, selection, and judgment to prompted LLMs; there is no formal guarantee that the output abstractions are causally faithful.
  • domain assumption RoBERTa mean-pooled embeddings of 'cause causes effect' templates capture semantic similarity between causal logics
    Section 3.2 computes distances in RoBERTa embedding space and clusters with complete linkage; if these embeddings do not align with causal similarity, the clusters and generated rules are arbitrary.
  • domain assumption Abstract rules transfer across corpora and surface lexicons
    The cross-dataset experiments (Section 6.4) assume rules induced from one corpus are valid for another; the paper tests this but the test is confounded by the data-leak issue.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Abstract Event Causal Rules: Induction and Application." pith.science (2026). https://pith.science/paper/GMPK46BJ

@misc{pith2026260805205,
  author       = {Pith},
  title        = {Pith review of: Abstract Event Causal Rules: Induction and Application},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GMPK46BJ}},
  note         = {Machine review of arXiv:2608.05205}
}
read the original abstract

Event-centric intelligent analytical systems heavily depend on explicit causal event knowledge for risk early warning, decision-making support and narrative comprehension. Nevertheless, existing instance-level causal pairs suffer severe generalization deficits on low-frequency long-tail and unseen event combinations. To address this limitation, this work proposes Abstract Event Causal Rule (AECR), a novel relation-level causal abstraction paradigm that transforms concrete cause-effect pairs into generalized abstract causal logic while retaining their intrinsic causal relationships. We design a multi-agent Concrete-to-Abstract Causal Induction (CACI) system coupled with similarity-constrained clustering to distill trustworthy AECRs from noisy raw causal data, based on which two complete AECR knowledge bases are built. To validate the practical utility of abstract causal knowledge, we propose an Abstract Rule-Guided Causal Attention Encoder (AR-GCAE), which injects the retrieved AECRs into the causality Graph Event Prediction (CGEP) benchmark task via rule-guided attention layers and gated representation fusion. Quantitative experimental results reveal that applying AECRs substantially strengthens the generalization capacity of event causal reasoning and brings consistent performance improvements to event prediction, with the most prominent gains observed on rare and unseen event samples.

Figures

Figures reproduced from arXiv: 2608.05205 by the authors.

Figure 1
Figure 1. Overview of the CACI multi-agent system for concrete event causal logic extraction. The five agents run sequentially [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Overview of AECR knowledge base and retriever. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Confusion matrix of human and machine evalua [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Overview of AR-GCAE Framework for Event Prediction. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Results of frequency-stratified evaluation on [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Results on low-resource scenario evaluation: Hit@1 [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: Results on the number of injected Top-𝐾 rules on the MAVEN-CGEP dataset. MAVEN-CGEP ESC-CGEP Model MRR Hit@1 Hit@10 MRR Hit@1 Hit@10 w/o AECR 34.8 27.1 51.5 19.3 15.2 27.5 w/o Gated Fusion 38.8 30.3 56.4 21.9 16.8 31.3 w Random Rules 35.0 27.1 51.6 19.7 15.0 28.5 Full …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 34 canonical work pages

  1. [1]

    Long Bai, Saiping Guan, Jiafeng Guo, Zixuan Li, Xiaolong Jin, and Xueqi Cheng

    work page

  2. [2]

    Pengfei Cao, Xinyu Zuo, Yubo Chen, Kang Liu, Jun Zhao, Yuguang Chen, and Weihua Peng. 2021. Knowledge-enriched event causality identification via latent structure induction networks. InProceedings of the 59th annual meeting of the association for computational linguistics and the 11th international joint conference on natural language processing (volume 1...

    work page 2021

  3. [3]

    Tommaso Caselli and Piek Vossen. 2017. The event storyline corpus: A new benchmark for causal and temporal relation extraction. InProceedings of the Events and Stories in the News Workshop. 77–86

    work page 2017

  4. [4]

    Nathanael Chambers and Dan Jurafsky. 2008. Unsupervised learning of narrative event chains. InProceedings of ACL-08: HLT. 789–797

    work page 2008

  5. [5]

    Xiao Ding, Kuo Liao, Ting Liu, Zhongyang Li, and Junwen Duan. 2019. Event representation learning enhanced with external commonsense knowledge. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Pro- cessing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 4894–4903

    work page 2019

  6. [6]

    Xinya Du, Zixuan Zhang, Sha Li, Pengfei Yu, Hongwei Wang, Tuan Lai, Xudong Lin, Ziqi Wang, Iris Liu, Ben Zhou, et al. 2022. RESIN-11: Schema-guided event prediction for 11 newsworthy scenarios. InProceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: System Demonstrat...

    work page 2022

  7. [7]

    Sharmishtha Dutta, Alex Gittens, Mohammed J Zaki, and Charu C Aggarwal

    work page

  8. [8]

    Mark Granroth-Wilding and Stephen Clark. 2016. What happens next? event prediction using a compositional neural network model. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 30

    work page 2016

Show all 43 references
  1. [9]

    Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Ab- hishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schel- ten, Alex Vaughan, et al . 2024. The llama 3 herd of models.arXiv preprint arXiv:2407.21783(2024)

    2024 arXiv

  2. [10]

    Saiping Guan, Xueqi Cheng, Long Bai, Fujun Zhang, Zixuan Li, Yutao Zeng, Xiaolong Jin, and Jiafeng Guo. 2023. What is Event Knowledge Graph: A Survey. IEEE Transactions on Knowledge and Data Engineering35, 7 (2023), 7569–7589. https://doi.org/10.1109/TKDE.2022.3180362

    2023

  3. [11]

    Stefan Heindorf, Yan Scholten, Henning Wachsmuth, Axel-Cyrille Ngonga Ngomo, and Martin Potthast. 2020. Causenet: Towards a causality graph extracted from the web. InProceedings of the 29th ACM international conference on information & knowledge management. 3023–3030

    2020

  4. [12]

    Jinhao Jiang, Kun Zhou, Zican Dong, Keming Ye, Xin Zhao, and Ji-Rong Wen

  5. [13]

    Cheng Jiayang, Lin Qiu, Chunkit Chan, Xin Liu, Yangqiu Song, and Zheng Zhang

  6. [14]

    Sha Li, Ruining Zhao, Manling Li, Heng Ji, Chris Callison-Burch, and Jiawei Han. 2023. Open-domain hierarchical event schema induction by incremental prompting and verification. InProceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1...

    2023

  7. [15]

    Zhongyang Li, Xiao Ding, and Ting Liu. 2018. Constructing narrative event evolutionary graph for script event prediction.arXiv preprint arXiv:1805.05081 (2018)

    2018 arXiv

  8. [16]

    Tian Liang, Zhiwei He, Wenxiang Jiao, Xing Wang, Yan Wang, Rui Wang, Yujiu Yang, Shuming Shi, and Zhaopeng Tu. 2024. Encouraging divergent thinking in large language models through multi-agent debate. InProceedings of the 2024 conference on empirical methods in natural languag...

    2024

  9. [17]

    Jian Liu, Yubo Chen, Kang Liu, Wei Bi, and Xiaojiang Liu. 2020. Event extraction as machine reading comprehension. InProceedings of the 2020 conference on empirical methods in natural language processing (EMNLP). 1641–1651

    2020

  10. [18]

    Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. Roberta: A robustly optimized bert pretraining approach.arXiv preprint arXiv:1907.11692 (2019)

    2019 arXiv

  11. [19]

    Linhao Luo, Yuan-Fang Li, Reza Haffari, and Shirui Pan. 2024. Reasoning on graphs: Faithful and interpretable large language model reasoning. InInterna- tional Conference on Learning Representations, Vol. 2024. 14400–14423

    2024

  12. [20]

    Shangwen Lv, Fuqing Zhu, and Songlin Hu. 2020. Integrating External Event Knowledge for Script Learning. InProceedings of the 28th International Confer- ence on Computational Linguistics, Donia Scott, Nuria Bel, and Chengqing Zong (Eds.). International Committee on Computation...

    2020 doi

  13. [21]

    Qing Lyu, Hongming Zhang, Elior Sulem, and Dan Roth. 2021. Zero-shot event extraction via transfer learning: Challenges and insights. InProceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natura...

    2021

  14. [22]

    Nasrin Mostafazadeh, Aditya Kalyanpur, Lori Moon, David Buchanan, Lauren Berkowitz, Or Biran, and Jennifer Chu-Carroll. 2020. GLUCOSE: GeneraLized and COntextualized story explanations. InProceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (E...

    2020

  15. [23]

    Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. 2022. Training language models to follow instructions with human feedback.Advances in neural information processing systems35 (...

    2022

  16. [24]

    Karl Pichotta and Raymond Mooney. 2016. Learning statistical scripts with LSTM recurrent neural networks. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 30

    2016

  17. [25]

    Nils Reimers and Iryna Gurevych. 2019. Sentence-bert: Sentence embeddings using siamese bert-networks. InProceedings of the 2019 conference on empirical methods in natural language processing and the 9th international joint conference on natural language processing (EMNLP-IJCN...

    2019

  18. [26]

    Angelika Romanou, Syrielle Montariol, Debjit Paul, Leo Laugier, Karl Aberer, and Antoine Bosselut. 2023. Crab: Assessing the strength of causal relationships between real-world events. InProceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 15...

    2023

  19. [27]

    Maarten Sap, Ronan Le Bras, Emily Allaway, Chandra Bhagavatula, Nicholas Lourie, Hannah Rashkin, Brendan Roof, Noah A Smith, and Yejin Choi. 2019. Atomic: An atlas of machine commonsense for if-then reasoning. InProceedings of the AAAI conference on artificial intelligence, Vo...

    2019

  20. [28]

    Weijia Shi, Sewon Min, Michihiro Yasunaga, Minjoon Seo, Richard James, Mike Lewis, Luke Zettlemoyer, and Wen-tau Yih. 2024. Replug: Retrieval-augmented black-box language models. InProceedings of the 2024 Conference of the North American Chapter of the Association for Computat...

    2024

  21. [29]

    Robyn Speer, Joshua Chin, and Catherine Havasi. 2017. Conceptnet 5.5: An open multilingual graph of general knowledge. InProceedings of the AAAI conference on artificial intelligence, Vol. 31

    2017

  22. [30]

    Zhengwei Tao, Zhi Jin, Xiaoying Bai, Haiyan Zhao, Chengfeng Dou, Yongqiang Zhao, Fang Wang, and Chongyang Tao. 2023. Seag: Structure-aware event causality generation. InFindings of the Association for Computational Linguistics: ACL 2023. 4631–4644

    2023

  23. [31]

    Vy Vo, Lizhen Qu, Tao Feng, Yuncheng Hua, Xiaoxi Kang, Songhai Fan, Tim Dwyer, Lay-Ki Soon, and Gholamreza Haffari. 2025. ACCESS: A Benchmark for Abstract Causal Event Discovery and Reasoning. InProceedings of the 2025 Con- ference of the Nations of the Americas Chapter of the...

    2025

  24. [32]

    Xiaozhi Wang, Yulin Chen, Ning Ding, Hao Peng, Zimu Wang, Yankai Lin, Xu Han, Lei Hou, Juanzi Li, Zhiyuan Liu, et al. 2022. Maven-ere: A unified large-scale dataset for event coreference, temporal, causal, and subevent relation extraction. InProceedings of the 2022 Conference ...

    2022

  25. [33]

    Yicong Wu, Guangyue Lu, Yuan Zuo, Huarong Zhang, and Junjie Wu. 2025. GRAPH-R1: Incentivizing the Zero-Shot Graph Learning Capability in LLMs via Explicit Reasoning. InProceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. 23920–23938

    2025

  26. [34]

    Matej Zečević, Moritz Willig, Devendra Singh Dhami, and Kristian Kersting

  27. [35]

    Chuanhong Zhan, Wei Xiang, Liang Chao, and Bang Wang. 2024. What would happen next? Predicting consequences from an event causality graph. InFindings of the Association for Computational Linguistics: EMNLP 2024. 820–832

    2024

  28. [36]

    Chuanhong Zhan, Wei Xiang, Chao Liang, and Bang Wang. 2025. Unified Se- mantic and Structural Modeling via Focal Attention for Causality Graph Event Prediction.IEEE Transactions on Audio, Speech and Language Processing(2025)

    2025

  29. [37]

    Ziwei Zheng, Qing Zhang, Chuanhong Zhan, and Bang Wang. 2026. TRACE: A Triplet-Based Robustness-Augmented Causal Encoder for Causality Graph Event Prediction. InICASSP 2026-2026 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, 19372–19376

    2026

  30. [38]

    Fangqi Zhu, Jun Gao, Changlong Yu, Wei Wang, Chen Xu, Xin Mu, Min Yang, and Ruifeng Xu. 2023. A generative approach for script event prediction via con- trastive fine-tuning. InProceedings of the AAAI conference on artificial intelligence, Vol. 37. 14056–14064. 13

    2023

  31. [39]

    Causal parrots: Large language models may talk causality but are not causal.arXiv preprint arXiv:2308.13067(2023)

    2023 arXiv

  32. [2021]

    InProceedings of the 2021 conference on empirical methods in natural language processing

    Integrating deep event-level and script-level information for script event prediction. InProceedings of the 2021 conference on empirical methods in natural language processing. 9869–9878

    2021

  33. [2023]

    InProceedings of the 2023 conference on empirical methods in natural language processing

    Structgpt: A general framework for large language model to reason over structured data. InProceedings of the 2023 conference on empirical methods in natural language processing. 9237–9251

    2023

  34. [2024]

    InProceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)

    Eventground: Narrative reasoning by grounding to eventuality-centric knowledge graphs. InProceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024). 6622–6642

    2024

  35. [2025]

    InProceedings of the AAAI Conference on Artificial Intelligence, Vol

    Replacing paths with connection-biased attention for knowledge graph completion. InProceedings of the AAAI Conference on Artificial Intelligence, Vol. 39. 14885–14892

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.