VaeDiff-DocRE: End-to-end Data Augmentation Framework for Document-level Relation Extraction

Khai Phan Tran; Wen Hua; Xue Li

REVIEW 4 major objections 6 minor 47 references

VaeDiff-DocRE: End-to-end Data Augmentation Framework for Document-level Relation Extraction

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper claims that VaeDiff-DocRE, a data-augmentation framework that generates synthetic entity-pair representations in embedding space using a VAE with a diffusion-model prior, outperforms existing DocRE methods on long-tail relations…

desk verdict A credible but modest embedding-space augmentation recipe for long-tail DocRE, with a wrong appendix derivation and an unresolved distribution-shift question; deserves review, needs revision. read the letter →

arxiv 2412.13503 v2 pith:DX555K73 submitted 2024-12-18 cs.CL cs.AI

Khai Phan Tran , Wen Hua , Xue Li This is my paper · ORCID

classification cs.CLcs.AI

keywords document-levelrelationextractiondataaugmentationlong-taildistributionvariationalautoencoderdiffusionprobabilisticmodelembeddingspacemulti-labelclassificationclassimbalance

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

Document-level relation extraction (DocRE) identifies relations between entity pairs across a whole document, and real datasets are heavily imbalanced: most pairs have no relation and a few relations cover most positive examples. This paper claims that the imbalance can be treated directly by generating extra training examples for underrepresented relations in the embedding space, rather than by redesigning loss functions that often trade minority-class gains against majority-class losses. It proposes VaeDiff-DocRE, which trains a variational autoencoder on entity-pair representations, uses a diffusion model as the VAE's latent prior so that generation can be conditioned on multiple relation labels at once, and mixes the resulting pseudo-features into the training batches of a retrained DocRE model. On Re-DocRED and DWIE, the framework reports higher F1 than prior state-of-the-art DocRE methods, with the largest gains on the long-tail F1 metric for rare relations. The point of the work, if correct, is that embedding-space oversampling with a learned relation-wise generative model is a practical, backbone-agnostic cure for long-tail DocRE.

What carries the argument

The load-bearing mechanism is the VaeDiff augmentation module, made of two parts. The first is EP-VAE, a variational autoencoder whose encoder projects entity-pair representations into a Gaussian latent space and whose decoder reconstructs them; the second is the Diffusion Prior, a denoising diffusion probabilistic model that replaces the Gaussian prior and is trained to predict the clean latent vector from a noised version. The diffusion model is conditioned through classifier-free guidance on a relation-label embedding computed as the sum of embeddings of all positive labels, which is what allows multi-label generation. The module is inserted into a three-stage hierarchical loop: stage 1 trains a baseline DocRE encoder and classifier, stage 2 trains VaeDiff on that encoder's entity-pair features, and stage 3 retrains the DocRE model while combining its own features with pseudo-features sampled from VaeDiff. The VAE supplies a compact, reconstructable representation space; the diffusion prior supplies a richer, class-conditioned latent distribution than a Gaussian; and the three-stage loop is what lets the augmentation adapt to the classifier being trained.

What would settle it

Re-run stage 3 on Re-DocRED with the stage-2 generative model left frozen but the document encoder re-initialized; if the reported gains in LTail F1 and overall F1 vanish or reverse, the augmentation depends on the original embedding space rather than on faithful oversampling, and measuring the distance between generated pseudo-features and real stage-3 features for the same relations would confirm whether the generator is sampling from a stale distribution.

Watch

Extended reading notes

Core claim

The central discovery, stated on the paper's own terms, is that the representations of entity pairs produced by a trained DocRE encoder organize into distinct relation-wise clusters, and that these clusters can be captured and sampled from. VaeDiff-DocRE builds an Entity-Pair VAE (EP-VAE) that reconstructs entity-pair representations from a compact latent space, then replaces the usual Gaussian prior on the latent variable with a denoising diffusion model. The diffusion model is class-conditioned on the sum of the embeddings of all positive relation labels for a pair, which lets one generated pseudo-feature encode several co-occurring relations. A three-stage training loop first trains a baseline DocRE model, then trains VaeDiff on the baseline's entity-pair representations, and finally retrains the DocRE model with the generated pseudo-features mixed into each batch. The authors report that this beats the compared methods on both benchmarks, for example by 1.6 and 2.27 F1 points on the DWIE development and test sets, and improves LTail F1 on Re-DocRED while keeping majority-relation F1 competitive.

Load-bearing premise

The load-bearing premise is that the stage-2 generative model, trained on representations from the stage-1 encoder, still produces useful pseudo-features after the stage-3 retraining changes the encoder's embedding space; the paper does not demonstrate that this transfer holds.

Editorial extensions

If this is right

Rare relations receive synthetic training examples in embedding space without new annotation, directly addressing positive-positive imbalance instead of re-weighting loss terms that trade majority-class accuracy for minority-class gains.
Because the stage-1 and stage-3 models can be any DocRE backbone, the augmentation module is portable: upgrading the underlying encoder or classifier should preserve the benefit.
Conditioning generation on the full multi-label vector means a single pseudo-entity-pair can carry several co-occurring relations, matching the multi-label structure of DocRE.
On the reported experiments, the framework raises overall F1 by 0.58 (BERT) and 0.68 (RoBERTa) on the Re-DocRED development set and by 2.27 F1 points on the DWIE test set, while improving LTail F1 on Re-DocRED test by 0.54 and 0.42 points over the state of the art.

Reading between the lines

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

A direct test of the framework's logic would freeze the stage-2 generative model while re-initializing the stage-3 encoder; if the gains collapse, the pseudo-features are tied to the stage-1 embedding geometry and the improvement is at least partly a regularizing effect rather than faithful oversampling.
A cheaper baseline worth testing is a conditional VAE with a plain Gaussian prior (or tuned Gaussian noise on real features); comparing against it would isolate what the diffusion prior contributes beyond added stochasticity.
The same embedding-space augmentation recipe could transfer to other extreme multi-label document tasks, such as event extraction or fine-grained entity typing, wherever a trained encoder yields label-conditioned clusters.

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.

Desk Editor's Note

read the letter

The paper is a serious attempt at embedding-space augmentation for long-tail DocRE, and the core combination is new: a VAE for relation-wise distributions with a diffusion prior to handle multi-label, wrapped in a three-stage training scheme. That's a clean idea, and the experiments are above average for the field — ten runs, significance tests, ablations, released code. The gains are modest but consistent, especially on LTail F1, and the ablation shows the full VaeDiff beats both Gaussian noise and no augmentation. Credit where it's due: the empirical work is reproducible and the central claim, that this helps long-tail relations without hurting common ones, is supported on two benchmarks.

The soft spots are real but fixable. First, the derivation of Eq. (9) in Appendix A.4 is wrong. Equation (32) drops the KL terms in an inequality that doesn't go through, so the theoretical justification as written is invalid. This doesn't necessarily kill the method — the loss still works empirically — but the paper needs to correct the math or remove the overclaim. Second, Algorithm 1 is ambiguous about labels: the pseudo-features are generated for positive entity pairs, but the loss is computed against the full label vector y on the concatenated set. The authors need to spell out how generated features are assigned labels. Third, the distribution-shift concern is legitimate but the stress-test overstates it. There is qualitative evidence in Figure 5 that the generated and encoded features overlap in the final model, so it's not true that there is 'no evidence.' What is missing is a quantitative check of whether the pseudo-features stay in-distribution during the early epochs of Stage 3, when the encoder is still drifting. Given the Gaussian-noise ablation already gains 0.13 F1, a skeptic can reasonably ask how much of the full gain is faithful oversampling versus structured regularization. That's a fair question for the authors to address, but it's a request for more analysis, not a demonstrated fatal flaw.

Overall, I'd send this to peer review. The method is novel, the experiments are solid, and the issues are presentation and analysis gaps rather than wrong results. A serious referee would want the derivation corrected, the label handling clarified, and a distribution-shift analysis added. It's a long-tail DocRE paper, so the audience is specialized, but within that audience it's worth the referee time.

Referee Report

4 major / 6 minor

Summary. The paper proposes VaeDiff-DocRE, a three-stage data augmentation framework for document-level relation extraction (DocRE). In Stage 1, a baseline DocRE model is trained to produce entity-pair representations that form relation-wise clusters. In Stage 2, a variational autoencoder (EP-VAE) is trained on these representations, with a conditional diffusion model serving as the latent prior (the 'VaeDiff' module). In Stage 3, the baseline DocRE model is retrained from scratch, and after a short warm-up the frozen VaeDiff generates pseudo entity-pair representations for positive pairs, which are mixed with the model's own features during training. Experiments on Re-DocRED and DWIE report consistent improvements over prior state-of-the-art methods, and ablations show that the augmentation contributes beyond simple Gaussian noise. The paper also releases code and reports average results over ten runs with significance tests.

Significance. If the reported improvements are robust and the augmentation mechanism works as claimed, the main contribution is a practical embedding-space augmentation method for long-tail DocRE, and it would be, to the authors' knowledge, the first application of diffusion probabilistic models to DocRE. The strengths are the reproducible experimental setup (two benchmarks, ten runs, significance tests, released code) and the clear isolation of the augmentation effect through the No Aug and Gaus Noise ablations. However, the central 'faithful oversampling' interpretation is currently not supported because of the distribution shift between the Stage 1 encoder used to train VaeDiff and the retrained Stage 3 encoder; the incremental gain over Gaussian noise could be a regularization effect rather than relation-wise oversampling. This distinction is important for the paper's claim to 'effectively address the long-tail distribution problem.'

major comments (4)

[Sections 2.3.2 and 2.3.3, Algorithm 1] The VaeDiff module is trained on entity-pair representations extracted from the Stage 1 encoder, but Stage 3 retrains the baseline DocRE model from scratch and starts augmentation after only n warmup epochs (n=5 for Re-DocRED, n=10 for DWIE, Section 3.1). No mechanism aligns the Stage 3 encoder's embedding space with the Stage 1 space, so the pseudo-features can become stale or out-of-distribution as the Stage 3 encoder drifts. The t-SNE visualization in Section 3.5 does not specify which encoder produced the 'encoded data', and it provides no quantitative distribution-distance measure. The Gaus Noise ablation in Table 4 shows that generic perturbation already improves F1 by 0.13 over KD-DocRE, while the full VaeDiff adds 0.45 more; this leaves open the possibility that the additional gain reflects a structured but stale feature distribution acting as a regularizer rather than faithful oversampling of minority relations. Please add a quantitative analysis of the distribution mismatch across Stage 3 training (e.g., MMD or centroid distances between generated and encoded features), or adapt the generator during Stage 3, or initialize Stage 3 from the Stage 1 encoder.
[Appendix A.4, Eq. (31)-(32) and Eq. (9)] The derivation of Eq. (9) is invalid as written. Eq. (31) is a valid lower bound on the ELBO in Eq. (30) after replacing the intractable log p_psi(z0) with the diffusion ELBO, but the displayed chain in Eq. (32) states that this lower bound is at most the original ELBO and at most E[log f_theta], which drops the KL terms and does not establish the claimed 'valid ELBO formulation'. The text should be rewritten to say that Eq. (31) is a lower bound that is maximized as a surrogate for the original ELBO, and that the regression objective L_Diff is then used as a further surrogate for the diffusion ELBO term. Additionally, the expectation in the last term of Eq. (9) should be over the forward diffusion process q(z_t | z_0), not merely over q_phi(z_p | p).
[Sections 2.2.2, Eq. (7), and Algorithm 1] The conditioning in the Diffusion Prior is the sum of the label embeddings of all ground-truth relations of an entity pair, and in Stage 3 the generator produces pseudo-features for whole positive pairs rather than for individual relations. The paper describes the modeling objects as relation-wise distributions (Section 2.2), but the trained distribution is actually over entity-pair representations conditioned on multi-label vectors. For a pair that expresses both head and tail relations (e.g., P580 with P582, as discussed in Section 3.5), the conditional signal is a mixture, so the generated feature does not specifically target the minority relation. To support the long-tail claim, the authors should either generate per-relation pseudo-features (conditioning on a single label and assigning that label to the generated instance) or analyze how often the augmented pairs are ones whose minority relations are not accompanied by majority relations.
[Sections 2.3.1 and 3.3, Table 1] The main comparison to KD-DocRE is confounded by the baseline loss change: Stage 1 replaces KD-DocRE's Adaptive Focal Loss with PMTEM loss plus Supervised Contrastive Learning (Section 2.3.1), so the F1 gains over KD-DocRE in Table 1 cannot be attributed solely to the augmentation module. The No Aug row in Table 4 controls for the loss and is weaker than KD-DocRE (75.20 vs. 75.31), which does show that the final gain over KD-DocRE requires augmentation; nevertheless, the paper should either report the same baseline with and without augmentation (i.e., KD-DocRE with PMTEM+SCL but no VaeDiff) or explicitly acknowledge that the reported SOTA comparison includes the loss change as a co-variate.

minor comments (6)

[Section 1 and Table 3] The text states that DWIE has 9 out of 66 relation labels in the long tail, but Table 3 lists 65 relations; please clarify the correct count.
[Section 3.5 and Figure 5] The caption of Figure 5 is missing the marker symbol for the encoded data; it says 'represents the encoded actual entity pair representations' but does not show which symbol (e.g., a dot) is used.
[Algorithm 1] The phrase 'Sample sampling' in line 23 of Algorithm 1 should be 'Sample noise', and the comment in lines 11-18 would be more readable if it explicitly stated that self-conditioning uses the previous estimate with stop-gradient.
[Section 3.1, Metrics] The acronym 'LT F1' appears in the description of Re-DocRED metrics; it should be 'LTail F1' for consistency with Tables 1 and 4.
[Section 3.1, Implementation details] The paper reports that all p-values are less than 0.05 following Dror et al. (2018), but it does not state which test was used (e.g., paired bootstrap or approximate randomization); please specify the test and the number of bootstrap samples.
[Limitation section] The Limitation section mentions longer GPU training and unverified specialized domains, but it does not mention the distribution-shift issue between the training of VaeDiff and its deployment in Stage 3; a sentence acknowledging this would align the stated limitations with the actual risks of the method.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: VaeDiff-DocRE's central claims are grounded in external held-out benchmarks; the label-conditioned augmentation loop is standard training design, not a reduction of the result.

full rationale

The paper's load-bearing assertion is an empirical F1 improvement over external SOTA methods on held-out Re-DocRED and DWIE dev/test sets (Section 3.3, Tables 1-2), so it cannot be equivalent by construction to the training inputs. The augmentation module (Section 2.2) is a VAE with a diffusion prior trained to reconstruct entity-pair representations from the Stage 1 encoder; the pseudo-features are then used to train the Stage 3 classifier. Although Algorithm 1 conditions generation on ground-truth label vectors and computes the classifier loss against those same labels, this is conditional data augmentation, not a fitted parameter renamed as a prediction; the reported generalization evidence is the held-out evaluation. The concern that the Stage 2 generative model may be stale after Stage 3 retraining (Sections 2.3.2-2.3.3) is a distribution-shift/validity question, not a circularity, because the benchmark numbers are not derived from the model's own outputs by definition. There are no load-bearing self-citations: KD-DocRE and other cited methods are external prior work. The Limitation section only notes longer training time and unexamined specialized domains, neither of which imports an unverified premise. No enumerated circularity pattern is present.

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

The method rests on standard VAE and diffusion machinery plus two domain assumptions: that relation-wise clusters exist in embedding space, and that the embedding space does not shift too much when the encoder is retrained in Stage 3. Hyperparameters m, w, warmup epochs, and conditioning probability are hand-tuned on development data and directly affect the reported results.

free parameters (4)

m = 2
Number of synthetic representations generated per positive entity pair in Stage 3 (Section 3.1). Directly controls augmentation strength and affects all reported F1 numbers; chosen on the development set.
w = 0.1
Classifier-free guidance weight in Eq. (6) balancing conditional and unconditional diffusion outputs; tuned on dev.
warmup_epochs n = 5 (Re-DocRED), 10 (DWIE)
Epochs before augmentation begins in Stage 3; per-dataset choice that determines how much the encoder has adapted before synthetic features are introduced.
self-condition/class-condition probability p = 0.1
Probability of applying self-conditioning and class-conditioning during diffusion training (Section 2.2.2).

assumptions (4)

standard math ELBO with reparameterization trick and the DDPM forward/reverse process equations (Ho et al. 2020) are valid and applicable.
Used to define EP-VAE training (Eq. 4) and diffusion prior (Eq. 5).
standard math Replacing the prior log-likelihood term in the VAE ELBO with the diffusion denoising objective gives a valid training loss (Wehenkel and Louppe 2021).
Core of Eq. (9); the paper's own derivation in Appendix A.4 contains an incorrect inequality, so this relies on the cited reference rather than the appendix.
domain assumption Entity pair representations for a given relation form distinct, learnable clusters in embedding space (Figure 2).
Motivates the whole augmentation idea; if the cluster structure is weak, generated samples will not be useful.
domain assumption The embedding distribution is stable enough between Stage 1 and Stage 3 that generated representations remain valid.
Needed for Stage 3 to work; not verified in the paper.

how reviews work

0 comments

Cite this review

Pith. "Pith review of VaeDiff-DocRE: End-to-end Data Augmentation Framework for Document-level Relation Extraction." pith.science (2026). https://pith.science/paper/DX555K73

@misc{pith2026241213503,
  author       = {Pith},
  title        = {Pith review of: VaeDiff-DocRE: End-to-end Data Augmentation Framework for Document-level Relation Extraction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DX555K73}},
  note         = {Machine review of arXiv:2412.13503}
}

read the original abstract

Document-level Relation Extraction (DocRE) aims to identify relationships between entity pairs within a document. However, most existing methods assume a uniform label distribution, resulting in suboptimal performance on real-world, imbalanced datasets. To tackle this challenge, we propose a novel data augmentation approach using generative models to enhance data from the embedding space. Our method leverages the Variational Autoencoder (VAE) architecture to capture all relation-wise distributions formed by entity pair representations and augment data for underrepresented relations. To better capture the multi-label nature of DocRE, we parameterize the VAE's latent space with a Diffusion Model. Additionally, we introduce a hierarchical training framework to integrate the proposed VAE-based augmentation module into DocRE systems. Experiments on two benchmark datasets demonstrate that our method outperforms state-of-the-art models, effectively addressing the long-tail distribution problem in DocRE.

Figures

Figures reproduced from arXiv: 2412.13503 by the authors.

**Figure 1.** Relation frequency in Re-DocRED (Tan et al., 2022b) and DWIE (Zaporojets et al., 2021) datasets. interest in exploring the problem of extracting relations across multiple sentences, a task known as Document-level Relation Extraction (DocRE) (Xu et al., 2021; Zhang et al., 2021; Zhou et al., 2021; Tan et al., 2022a). Notably, the distribution of relations in datasets for the DocRE task often exhibits significant imb… view at source ↗

**Figure 2.** Relation-wise distribution from Re-DocRED [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 3.** Structure of EP-VAE module. 2.2.1 Entity Pair VAE The Entity Pair VAE, referred to as EP-VAE, consists of an encoder and a decoder fθ(·), parameterized by θ. The encoder includes two components: (1) a projector qϕ(·) that maps the input to the EPVAE latent space, and (2) separate projectors for the mean (µϕ) and log variance (σϕ), which estimate the parameters of the posterior distribution in the latent space. B… view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Overview of our VaeDiff-DocRE framework. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

**Figure 5.** Figure 5: t-SNE visualization of relation-wise distri [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 30 canonical work pages

[1]

online" 'onlinestring :=

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

Ting Chen, Ruixiang ZHANG, and Geoffrey Hinton. 2023. https://openreview.net/forum?id=3itjR9QxFw Analog bits: Generating discrete data using diffusion models with self-conditioning . In The Eleventh International Conference on Learning Representations

work page 2023
[4]

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT : Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 Conference of the North A merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) , pages 4171--4186, M...

2019
[5]

Rotem Dror, Gili Baumer, Segev Shlomov, and Roi Reichart. 2018. The hitchhiker’s guide to testing statistical significance in natural language processing. In Proceedings of the 56th annual meeting of the association for computational linguistics (volume 1: Long papers), pages 1383--1392

2018
[6]

Markus Eberts and Adrian Ulges. 2021. An end-to-end model for entity-level relation extraction using multi-instance learning. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 3650--3660

work page 2021
[7]

Jia Guo, Stanley Kok, and Lidong Bing. 2023. Towards integration of discriminability and robustness for document-level relation extraction. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pages 2606--2617

work page 2023
[8]

Ridong Han, Tao Peng, Benyou Wang, Lu Liu, Prayag Tiwari, and Xiang Wan. 2024. Document-level relation extraction with relation correlations. Neural Networks, 171:14--24

work page 2024

Show all 47 references

[9]

Jonathan Ho, Ajay Jain, and Pieter Abbeel. 2020. Denoising diffusion probabilistic models. Advances in neural information processing systems, 33:6840--6851

2020
[10]

Jonathan Ho, Nal Kalchbrenner, Dirk Weissenborn, and Tim Salimans. 2019. Axial attention in multidimensional transformers. arXiv preprint arXiv:1912.12180

2019 arXiv
[11]

Jonathan Ho and Tim Salimans. 2021. https://openreview.net/forum?id=qw8AKxfYbI Classifier-free diffusion guidance . In NeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications

2021
[12]

Sergey Ioffe and Christian Szegedy. 2015. Batch normalization: Accelerating deep network training by reducing internal covariate shift. In Proceedings of the 32nd International Conference on Machine Learning, volume 37, pages 448--456. PMLR

2015
[13]

Feng Jiang, Jianwei Niu, Shasha Mo, and Shengda Fan. 2022. Key mention pairs guided document-level relation extraction. In Proceedings of the 29th International Conference on Computational Linguistics, pages 1904--1914

2022
[14]

Prannay Khosla, Piotr Teterwak, Chen Wang, Aaron Sarna, Yonglong Tian, Phillip Isola, Aaron Maschinot, Ce Liu, and Dilip Krishnan. 2020. Supervised contrastive learning. Advances in neural information processing systems, 33:18661--18673

2020
[15]

Diederik Kingma and Max Welling. 2013. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114

2013 arXiv
[16]

Junpeng Li, Zixia Jia, and Zilong Zheng. 2023. Semi-automatic data enhancement for document-level relation extraction with distant supervision from large language models. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 5495--5505

2023
[17]

Xingzuo Li, Kehai Chen, Yunfei Long, and Min Zhang. 2024. Llm with relation classifier for document-level relation extraction. arXiv preprint arXiv:2408.13889

2024 arXiv
[18]

Tsung-Yi Lin, Priya Goyal, Ross Girshick, Kaiming He, and Piotr Doll \'a r. 2017. Focal loss for dense object detection. In Proceedings of the IEEE international conference on computer vision, pages 2980--2988

2017
[19]

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 arXiv
[20]

Ilya Loshchilov and Frank Hutter. 2019. https://openreview.net/forum?id=Bkg6RiCqY7 Decoupled weight decay regularization . In International Conference on Learning Representations

2019
[21]

Ilya Loshchilov and Frank Hutter. 2022. Sgdr: Stochastic gradient descent with warm restarts. In International Conference on Learning Representations

2022
[22]

Diego Marcheggiani and Ivan Titov. 2016. Discrete-state variational autoencoders for joint discovery and factorization of relations. Transactions of the Association for Computational Linguistics, 4:231--244

2016
[23]

Guoshun Nan, Zhijiang Guo, Ivan Sekuli \'c , and Wei Lu. 2020. Reasoning with latent structure refinement for document-level relation extraction. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 1546--1557

2020
[24]

Alexander Quinn Nichol and Prafulla Dhariwal. 2021. Improved denoising diffusion probabilistic models. In International conference on machine learning, pages 8162--8171. PMLR

2021
[25]

Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, et al. 2019. Pytorch: An imperative style, high-performance deep learning library. Advances in neural information processing systems, 32

2019
[26]

Qi Sun, Kun Huang, Xiaocui Yang, Rong Tong, Kun Zhang, and Soujanya Poria. 2024. Consistency guided knowledge retrieval and denoising in llms for zero-shot document-level relation triplet extraction. In Proceedings of the ACM on Web Conference 2024, pages 4407--4416

2024
[27]

Qingyu Tan, Ruidan He, Lidong Bing, and Hwee Tou Ng. 2022 a . Document-level relation extraction with adaptive focal loss and knowledge distillation. In Findings of the Association for Computational Linguistics: ACL 2022, pages 1672--1681

2022
[28]

Qingyu Tan, Lu Xu, Lidong Bing, and Hwee Tou Ng. 2023. Class-adaptive self-training for relation extraction with incompletely annotated training data. In Findings of the Association for Computational Linguistics: ACL 2023, pages 8630--8643

2023
[29]

Qingyu Tan, Lu Xu, Lidong Bing, Hwee Tou Ng, and Sharifah Mahani Aljunied. 2022 b . Revisiting docred-addressing the false negative problem in relation extraction. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 8472--8487

2022
[30]

Laurens Van der Maaten and Geoffrey Hinton. 2008. Visualizing data using t-sne. Journal of machine learning research, 9(11)

2008
[31]

Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, ukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Advances in Neural Information Processing Systems, volume 30

2017
[32]

Jize Wang, Xinyi Le, Xiaodi Peng, and Cailian Chen. 2023. Adaptive hinge balance loss for document-level relation extraction. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 3872--3878

2023
[33]

Antoine Wehenkel and Gilles Louppe. 2021. Diffusion priors in variational autoencoders. In ICML Workshop on Invertible Neural Networks, Normalizing Flows, and Explicit Likelihood Models

2021
[34]

Thomas Wolf, Lysandre Debut, Victor Sanh, Julien Chaumond, Clement Delangue, Anthony Moi, Pierric Cistac, Tim Rault, Remi Louf, Morgan Funtowicz, Joe Davison, Sam Shleifer, Patrick von Platen, Clara Ma, Yacine Jernite, Julien Plu, Canwen Xu, Teven Le Scao, Sylvain Gugger, Mari...

2020
[35]

Ruibin Xiong, Yunchang Yang, Di He, Kai Zheng, Shuxin Zheng, Chen Xing, Huishuai Zhang, Yanyan Lan, Liwei Wang, and Tieyan Liu. 2020. On layer normalization in the transformer architecture. In International Conference on Machine Learning, pages 10524--10533. PMLR

2020
[36]

Kun Xu, Siva Reddy, Yansong Feng, Songfang Huang, and Dongyan Zhao. 2016. Question answering on freebase via relation extraction and textual evidence. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 2326--2336

2016
[37]

Wang Xu, Kehai Chen, and Tiejun Zhao. 2021. Document-level relation extraction with reconstruction. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pages 14167--14175

2021
[38]

Lilong Xue, Dan Zhang, Yuxiao Dong, and Jie Tang. 2024. Autore: document-level relation extraction with large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations), pages 211--220

2024
[39]

Yuan Yao, Deming Ye, Peng Li, Xu Han, Yankai Lin, Zhenghao Liu, Zhiyuan Liu, Lixin Huang, Jie Zhou, and Maosong Sun. 2019. Docred: A large-scale document-level relation extraction dataset. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguisti...

2019
[40]

Chenhan Yuan and Hoda Eldardiry. 2021. Unsupervised relation extraction: A variational autoencoder approach. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 1929--1938

2021
[41]

Klim Zaporojets, Johannes Deleu, Chris Develder, and Thomas Demeester. 2021. Dwie: An entity-centric dataset for multi-task document-level information extraction. Information Processing & Management, 58(4):102563

2021
[42]

Shuang Zeng, Runxin Xu, Baobao Chang, and Lei Li. 2020. Double graph based reasoning for document-level relation extraction. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 1630--1640

2020
[43]

Ningyu Zhang, Xiang Chen, Xin Xie, Shumin Deng, Chuanqi Tan, Mosha Chen, Fei Huang, Luo Si, and Huajun Chen. 2021. Document-level relation extraction as semantic segmentation. In Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence. Internatio...

2021
[44]

Yifan Zhang, Bingyi Kang, Bryan Hooi, Shuicheng Yan, and Jiashi Feng. 2023. Deep long-tailed learning: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(9):10795--10816

2023
[45]

Yuhao Zhang, Victor Zhong, Danqi Chen, Gabor Angeli, and Christopher D Manning. 2017. Position-aware attention and supervised data improve slot filling. In Conference on Empirical Methods in Natural Language Processing

2017
[46]

Wenxuan Zhou, Kevin Huang, Tengyu Ma, and Jing Huang. 2021. Document-level relation extraction with adaptive thresholding and localized context pooling. In Proceedings of the AAAI conference on artificial intelligence, volume 35, pages 14612--14620

2021
[47]

Yang Zhou and Wee Sun Lee. 2022. None class ranking loss for document-level relation extraction. In Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, IJCAI-22 , pages 4538--4544. International Joint Conferences on Artificial Intelligenc...

2022

Pith tools

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

[1] [1]

online" 'onlinestring :=

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

[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]

Ting Chen, Ruixiang ZHANG, and Geoffrey Hinton. 2023. https://openreview.net/forum?id=3itjR9QxFw Analog bits: Generating discrete data using diffusion models with self-conditioning . In The Eleventh International Conference on Learning Representations

work page 2023

[4] [4]

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT : Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 Conference of the North A merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) , pages 4171--4186, M...

2019

[5] [5]

Rotem Dror, Gili Baumer, Segev Shlomov, and Roi Reichart. 2018. The hitchhiker’s guide to testing statistical significance in natural language processing. In Proceedings of the 56th annual meeting of the association for computational linguistics (volume 1: Long papers), pages 1383--1392

2018

[6] [6]

Markus Eberts and Adrian Ulges. 2021. An end-to-end model for entity-level relation extraction using multi-instance learning. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 3650--3660

work page 2021

[7] [7]

Jia Guo, Stanley Kok, and Lidong Bing. 2023. Towards integration of discriminability and robustness for document-level relation extraction. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pages 2606--2617

work page 2023

[8] [8]

Ridong Han, Tao Peng, Benyou Wang, Lu Liu, Prayag Tiwari, and Xiang Wan. 2024. Document-level relation extraction with relation correlations. Neural Networks, 171:14--24

work page 2024