Pith. sign in

REVIEW 5 major objections 6 minor 1 cited by

MLDGG: Meta-Learning for Domain Generalization on Graphs

T0 review · 5 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read MLDGG claims that node-level domain generalization on graphs is best achieved by meta-learning a shared initialization for a structure learner that rewires task-irrelevant edges and a representation learner that disentangles…

desk verdict Sensible MAML+disentanglement recipe with modest same-dataset gains, but the cross-dataset headline results are unfair comparisons over misaligned label spaces. read the letter →

arxiv 2411.12913 v1 pith:MC3EBMNB submitted 2024-11-19 cs.LG cs.AI

Qin Tian , Chen Zhao , Minglai Shao , Wenjun Wang , Yujie Lin , Dong Li This is my paper · ORCID
classification cs.LGcs.AI
keywords domaingeneralizationgraphneuralnetworksmeta-learningstructurelearningdisentangledrepresentationnodeclassificationdistributionshiftMAML
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper sets out to show that node-level domain generalization on graphs can be improved by treating each source graph as a meta-learning task and learning a shared initialization for two components: a structure learner that rewires task-irrelevant edges and a representation learner that separates domain-invariant semantic information from domain-specific variation. The motivation is that existing graph domain-generalization methods learn static encoders and therefore cannot adapt when the target graph's topology or features differ sharply from the source. MLDGG trains the two learners with a MAML-style bi-level loop, so that at deployment the learned parameters can be fine-tuned on a small support set from the unseen target graph. Experiments on TWITCH, Facebook-100, and WebKB under same-dataset, cross-dataset, and mixed-source settings report consistent gains over invariant-learning, data-augmentation, and meta-learning baselines. The transfer result rests on a causal invariance assumption: the generative mechanisms $p(r|s,v)$ and $p(y|s)$ stay fixed across domains, and only the prior $p(s,v)$ changes.

What carries the argument

The load-bearing object is the meta-learned initialization $\theta=\{\theta_t,\theta_r\}$, where $\theta_t$ parameterizes the structure learner and $\theta_r$ parameterizes the semantic encoder, variation encoder, and decoder of the representation learner. The structure learner turns raw adjacency into a refined graph $A'$ by sampling from a trainable similarity matrix and is trained by policy gradient on a smoothness/sparsity reward; this is the component that suppresses task-unrelated edges and is meant to transfer shared topology patterns. The representation learner uses a variational ELBO, $L_{ELBO}$, derived from the factorization $p(s,v,r,y)=p(s,v)p(r|s,v)p(y|s)$, with $q(y|r)=\mathbb{E}_{q(s,v|r)}[p(y|s)]$ linking the latent factors to the label; its three terms enforce label prediction from $s$, reconstruction of $r$ from $(s,v)$, and closeness of the variational posterior to the prior. The meta-loop alternates an inner update $\theta'^i=\theta-l_{in}\nabla L^i_{sup}$ per source graph with an outer update $\theta=\theta-l_{out}\nabla\frac{1}{M}\sum_i L^i_{qry}$, which is what makes the initialization adaptable to an unseen target graph.

What would settle it

Build a synthetic pair of graph domains with a known generative model where the label is determined by $s$ in the source graphs but by $v$ in the target graph, while keeping $p(r|s,v)$ fixed. If MLDGG fine-tuned on a small target support set still beats a plain GCN trained on the same support, then the claimed semantic/variation disentanglement is not the mechanism behind the result; if its advantage disappears, the causal invariance assumption is confirmed as the load-bearing premise.

Watch

Extended reading notes

Core claim

The central claim is that a MAML-style meta-initialization, applied jointly to a structure learner and a representation learner, yields graph models that transfer across domains better than static encoders. Concretely, the structure learner $f_t$ computes a similarity matrix $F_{jk}=\delta(r_j\odot\hat{w}, r_k\odot\hat{w})$, samples a refined adjacency $A'$ from Bernoulli factors, and optimizes the sampling policy with the REINFORCE gradient using a smoothness-plus-sparsity reward. The representation learner $f_r$ factorizes the GNN output $r$ into a semantic factor $s$ and a variation factor $v$ through a variational ELBO built on the causal generative model $p(s,v,r,y)=p(s,v)p(r|s,v)p(y|s)$; the classifier uses only $s$, while $s$ and $v$ together reconstruct $r$. The two learners are updated in an inner loop on each source graph's support set and an outer meta-loop on its query set, yielding initial parameters $\theta=\{\theta_t,\theta_r\}$ that adapt to a target graph by a few fine-tuning steps. The empirical section reports that this recipe outperforms state-of-the-art baselines across three distribution-shift settings, with the independent-factors variant MLDGG-ind performing best.

Load-bearing premise

The claim collapses if the generative mechanism that maps latent factors to node representations and labels is not the same across domains; the paper assumes only the mixing of semantic and variation factors changes, so a target domain that changes how representations or labels are produced would break the transfer.

Editorial extensions

If this is right

  • Target graphs never need to appear during training: the learned initialization is fine-tuned on a support set from the target graph, and the sensitivity analysis shows that cross-dataset targets adapt with very few gradient steps.
  • Removing the structure learner costs about 2-3% accuracy and removing the representation learner costs 3-6%, so both components are load-bearing for the reported gains.
  • Forcing the semantic and variation factors to be independent (MLDGG-ind) generally improves accuracy over the correlated version, implying that domain-specific variation actively hurts generalization when entangled with semantics.
  • The upper bound in Theorem 1 says that target error is controlled by source risk plus two JS-distance terms, one measuring how much semantic-conditioned distributions vary across source domains and one measuring the gap between target and source joint distributions; perfect disentanglement would drive the semantic-conditional term to zero.

Reading between the lines

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

  • The paper's invariance premise is the binding constraint: if a real target domain changes $p(y|s)$ or $p(r|s,v)$ rather than just the prior $p(s,v)$, the semantic code learned from source graphs may carry no label signal, and the reported advantage should shrink; a synthetic dataset with a known generative process could test this directly.
  • Although experiments are node-level, the same structure learner operates on general adjacency matrices, so the recipe may extend to graph- and edge-level generalization where topology shifts are even more pronounced.
  • The REINFORCE sampling of $A'$ with $H$ sampled structures per forward pass is a computational bottleneck; a Gumbel-style continuous relaxation or a learned edge-pruning head could make the approach scale to larger graphs, but the paper does not explore that direction.
  • The appendix itself flags that Theorem 1's printed $\min$ term is not rigorous and replaces it with a $\max$, so the stated bound should be read in its corrected, looser form rather than as first printed.
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 / 6 minor

Summary. The paper proposes MLDGG, a meta-learning framework for node-level domain generalization on graphs. It combines a structure learner that rewires graphs via a learned similarity matrix and a representation learner that disentangles node representations into domain-invariant semantic factors and domain-specific variation factors using a variational objective. The two learners are integrated via MAML-style bi-level optimization to obtain initializations that can be fine-tuned on target graphs. The authors report experiments on TWITCH, FB-100, and WEBKB under three cross-domain settings, claiming consistent improvements over existing graph domain generalization baselines.

Significance. The central idea—using meta-learned initializations for both structure learning and semantic/variation disentanglement to enable fast adaptation in graph domain generalization—is timely and potentially useful, especially as most prior work relies on static encoders. The same-dataset experiments (S1T1) show modest but consistent gains over GraphGlow and MD-Gram, and the ablation study indicates each component contributes. However, the significance is severely limited by the empirical protocol for cross-dataset settings and by the unsupported theoretical analysis. The theoretical bounds are not machine-checked and contain unproved lemmas and algebraic errors, so they do not provide a reliable foundation for the method. The reported cross-dataset gains appear to conflate domain generalization with few-shot task adaptation, which changes the claim that the paper actually establishes.

major comments (5)
  1. [Section 6.1.3, Tables 3-6] The cross-dataset settings do not share a label space between source and target. TWITCH-EXPLICIT and FB-100 are binary with semantically different prediction targets, and WEBKB has five web-page categories; 'label expansion' in Section 6.1.3 only standardizes label dimensions, not label semantics. Consequently, a model trained on TWITCH and evaluated on FB-100 is predicting an unrelated property. The non-adaptive baselines score near or below chance (e.g., Table 3: EERM 13.1%, FLOOD 13.1%, ERM 12.3%; Table 4: EERM 17.0%, ERM 15.9%), which indicates that the evaluation protocol misaligns class identities. This casts doubt on whether Tables 3-6 measure domain generalization at all.
  2. [Sections 4.3 and 6.1.4, Tables 3-6] MLDGG is fine-tuned on a labeled target support set during meta-testing (Section 4.3, Eq. (12)-(13); Algorithm 1 uses T_eT_sup; Section 6.1.4 specifes 1 to 40 update steps), while the non-meta baselines (ERM, EERM, FLOOD, SRGNN, Mixup) are evaluated without any target-domain adaptation. The large gains in cross-dataset tables (e.g., 12-32 percentage points over ERM) may therefore be due to the ability to relearn the target task from the support set rather than to a domain-generalizing prior. To support the claimed comparison, all baselines should be fine-tuned on the same labeled target support set under the same protocol, or the paper should clearly reframe the problem as few-shot cross-task adaptation.
  3. [Section 5 and Appendix E, Lemma 2] The proof of Theorem 1 relies on Lemma 2, which states an exact additive decomposition of the JS distance: dJS(Pei S,Y, Pej S,Y) = dJS(Pei Y, Pej Y) + sqrt(2 E_y[dJS(Pei S|Y, Pej S|Y)^2]). This identity does not follow from standard properties of JS distance (there is no chain rule for JS divergence analogous to the KL chain rule), and no proof is given in Appendix E. The appendix also contains an explicit note that a 'min' was replaced by 'max' in the triangle inequality step, indicating that the bound derivation is not finalized. Without Lemma 2, the upper bound in Eq. (14) does not close.
  4. [Section 4.2, Eq. (10) and Appendix B] The ELBO algebra in Eq. (10) does not close. With q(y|r) defined as E_{q(s,v|r)}[p(y|s)], the first term evaluates to log q(y|r), not to the negative cross-entropy loss as stated. The derivation in Eq. (17) of Appendix B mishandles the expectation with respect to q(s,v|r)p(y|s), and placing 1/q(y|r) outside the expectation is not justified. As a result, the claimed decomposition into cross-entropy, reconstruction, and regularization terms is unsupported, and the objective L in Eq. (11) is not established as a valid variational lower bound.
  5. [Section 4.1, Eq. (4)] The smoothness regularization term in Eq. (4) is written as -alpha * sum_{j,k} A'_{j,k} ||r_j - r_j||_2^2, which is identically zero because it compares each node to itself. As printed, the reward function B in Eq. (4) reduces to only the sparsity term -beta||A'||_0, making the smoothness component vacuous. The intended term, presumably ||r_j - r_k||_2^2, is essential for the structure learner to encourage smooth node representations; the error must be corrected for Eq. (4)-(6) to describe the method accurately.
minor comments (6)
  1. [Abstract and Section 1] The phrase 'ensuring effective performance on the testing set despite disparities' is vague; consider defining the goal as minimizing expected error on an unseen target domain.
  2. [Section 3, Notations] The symbol 'E' is used both for the set of domains and for expectation, and 'Es' denotes both the semantic encoder and the set of source domains; this creates confusion in Section 4 and Appendix B.
  3. [Section 4.1, Eq. (3)] The similarity function delta(r_j ⊙ w, r_k ⊙ w) is not specified beyond 'simple dot-product and so on'; also, the weight vector w has dimension m but the node representation r has dimension d, and no constraint is stated that m = d.
  4. [Section 6.1.3] The sentence 'we employ zero-padding for feature dimensions and label expansion to standardize them after comparing different padding methods' provides no detail on the padding methods compared or the final procedure; this matters for reproducibility.
  5. [Theorem 2] In Eq. (15), the condition uses dJS(Pei Y, PeT Y) >= dJS(Pei S, PeT S), but the proof applies the bound to dJS(Pei A,X, PeT A,X). Since Lemma 1 only gives dJS(Pei S, PeT S) <= dJS(Pei A,X, PeT A,X), the stated condition does not imply the necessary inequality, so the lower bound may be vacuous.
  6. [Tables and Figures] There are typos: 'Dose MLDGG surpass the the state-of-the-art' in Section 6, 'Avg' appears inconsistently, Table 6 has the column header 'Amherst41', and Fig. 5 says 'T-sne' rather than 't-SNE'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: MLDGG's empirical claims are benchmarked against external baselines, and the causal disentanglement is an explicit modeling assumption rather than a construction that reproduces the claimed result.

full rationale

The central empirical claim, that MLDGG consistently outperforms state-of-the-art baselines for node-level prediction across three distribution-shift settings (Tables 2-6), is obtained by optimizing the meta-learning objective in Eq. (11) with inner/outer updates (Eqs. 12-13) and comparing against external methods implemented from their published code. The causal generative assumption in Section 4.2, that p(r|s,v) and p(y|s) are invariant across domains while p(s,v) changes, is an explicit modeling premise used to motivate the representation learner; it is not derived from the ELBO and does not by itself guarantee the reported accuracy gains. The ELBO in Eq. (10) includes a classification term, but that term is a standard supervised signal on source and support data, not a relabeling of the target test metric as a training objective. The theoretical bounds in Theorems 1-2 are generic JS-distance inequalities and are not used as evidence that the specific architecture must win; notably, the appendix even records a correction to the proof's min/max usage, indicating the analysis is not being tailored to force a predetermined conclusion. Self-citations appear only in the related-work context and are not load-bearing for the method's derivation. The cross-dataset label-alignment and target-support fine-tuning concerns raised by skeptics are experimental comparability questions, not cases where a prediction reduces by construction to its own inputs. Accordingly, no circular step is present.

Assumptions & free parameters 7 free parameters · 5 assumptions · 2 invented entities

The framework introduces learned structure weights, several loss weights with no reported tuning protocol, and latent variables s and v that are not externally validated. The transfer claim rests on a causal invariance assumption and an unproved JS-decomposition lemma. No code or data processing pipeline is provided to audit these choices.

free parameters (7)
  • similarity weight vector ŵ = trainable, dimension m not specified
    Eq. (3) uses ŵ to compute edge weights Fjk; this is the learned metric of the structure learner.
  • lambda (λ) = tuned in [0,1]
    Eq. (7) mixes GNN output on original and learned graphs; sensitivity analysis in Fig. 7 shows performance varies with λ.
  • alpha (α) = tuned in [0,1]
    Eq. (4) uses α as the smoothness reward weight; no tuning protocol is reported.
  • beta (β) = tuned in [0,1]
    Eq. (4) uses β as the sparsity reward weight; no tuning protocol is reported.
  • lambda_r (λr) = not reported
    Eq. (11) combines ELBO and structure regularization loss; the value is not stated.
  • latent dimensions of s and v = not specified
    Sec. 4.2.1 defines Gaussian posteriors and priors but does not state the dimensionality of the semantic and variation factors.
  • inner and outer learning rates = lin=1e-3, lout=1e-1
    Hyperparameters chosen for the MAML update; sensitivity is only reported for gradient steps, not rates.
assumptions (5)
  • domain assumption p(r|s,v) and p(y|s) are invariant across domains, and change of prior p(s,v) is the only source of domain change
    Sec. 4.2 states this causal invariance premise; the representation learner is built on it and it is not validated on the datasets.
  • domain assumption q(s,v|r) is Gaussian with diagonal covariance and p(s,v) is a multivariate Gaussian parameterized by Cholesky decomposition
    Sec. 4.2.1 chooses the variational family; if misspecified, maximizing the ELBO may not produce the intended disentanglement.
  • ad hoc to paper Lemma 2: dJS(P_S,Y; P_S',Y) = dJS(P_Y; P_Y') + sqrt(2 E_Y[dJS(P_S|Y; P_S'|Y)^2])
    Invoked in Appendix E.1 to derive Theorem 1 but stated without proof; this exact additive decomposition of JS distance is not a standard identity.
  • domain assumption MAML core assumption that some internal representations are better suited to transfer learning
    Sec. 3 adopts the standard MAML premise; the meta-objective only helps if source graph tasks share structure with the target graph task.
  • standard math REINFORCE gradient estimator is unbiased under Bernoulli sampling with distribution Φ(A')
    Eqs. (5)-(6) use a standard policy gradient estimator, though Eq. (4) as printed contains a typo that makes the smoothness reward term zero.
invented entities (2)
  • domain-invariant semantic factor s
    purpose: Latent variable intended to determine the label and transfer across domains.
    Sec. 4.2 introduces s as an internal construct; only t-SNE visualization is provided, with no external falsifiable prediction.
  • domain-specific variation factor v
    purpose: Latent variable intended to absorb label-independent, domain-specific information and aid reconstruction.
    Sec. 4.2 introduces v as an internal construct; its independence from labels is not externally validated.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MLDGG: Meta-Learning for Domain Generalization on Graphs." pith.science (2026). https://pith.science/paper/MC3EBMNB

@misc{pith2026241112913,
  author       = {Pith},
  title        = {Pith review of: MLDGG: Meta-Learning for Domain Generalization on Graphs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MC3EBMNB}},
  note         = {Machine review of arXiv:2411.12913}
}
read the original abstract

Domain generalization on graphs aims to develop models with robust generalization capabilities, ensuring effective performance on the testing set despite disparities between testing and training distributions. However, existing methods often rely on static encoders directly applied to the target domain, constraining its flexible adaptability. In contrast to conventional methodologies, which concentrate on developing specific generalized models, our framework, MLDGG, endeavors to achieve adaptable generalization across diverse domains by integrating cross-multi-domain meta-learning with structure learning and semantic identification. Initially, it introduces a generalized structure learner to mitigate the adverse effects of task-unrelated edges, enhancing the comprehensiveness of representations learned by Graph Neural Networks (GNNs) while capturing shared structural information across domains. Subsequently, a representation learner is designed to disentangle domain-invariant semantic and domain-specific variation information in node embedding by leveraging causal reasoning for semantic identification, further enhancing generalization. In the context of meta-learning, meta-parameters for both learners are optimized to facilitate knowledge transfer and enable effective adaptation to graphs through fine-tuning within the target domains, where target graphs are inaccessible during training. Our empirical results demonstrate that MLDGG surpasses baseline methods, showcasing its effectiveness in three different distribution shift settings.

Figures

Figures reproduced from arXiv: 2411.12913 by the authors.

Figure 1
Figure 1. Visualizations of distribution shifts on graphs demonstrated using energy scores [ [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. An overview of MLDGG. Each source graph is viewed as a task. For each task, the parameters [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Ablation study for MLDGG under three distinct cross-domain scenarios. [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Ablation study for MLDGG-ind under three distinct cross-domain scenarios. [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: The demonstration of the effectiveness of representation learner on [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Accuracy with different gradient steps during [PITH_FULL_IMAGE:figures/full_fig_p025_6.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. Out-of-Distribution Detection in Heterogeneous Graphs via Energy Propagation

    cs.LG 2025-04 conditional novelty 5.0 of 10

    OODHG detects out-of-distribution nodes in heterogeneous graphs by propagating energy scores along meta-paths and classifying the remaining in-distribution nodes.

Reference graph

Works this paper leans on

66 extracted references · 42 canonical work pages · cited by 1 Pith paper

  1. [1]

    Domain generalization via invariant feature representation

    Krikamol Muandet, David Balduzzi, and Bernhard Schölkopf. Domain generalization via invariant feature representation. In International conference on machine learning, pages 10–18. PMLR, 2013

    2013

  2. [2]

    Domain generalization with adversarial feature learning

    Haoliang Li, Sinno Jialin Pan, Shiqi Wang, and Alex C Kot. Domain generalization with adversarial feature learning. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 5400–5409, 2018

    2018

  3. [3]

    Supervised algorithmic fairness in distribution shifts: A survey

    Minglai Shao, Dong Li, Chen Zhao, Xintao Wu, Yujie Lin, and Qin Tian. Supervised algorithmic fairness in distribution shifts: A survey. arXiv preprint arXiv:2402.01327, 2024

    arXiv 2024

  4. [4]

    Algorithmic fairness generalization under covariate and dependence shifts simultaneously

    Chen Zhao, Kai Jiang, Xintao Wu, Haoliang Wang, Latifur Khan, Christan Grant, and Feng Chen. Algorithmic fairness generalization under covariate and dependence shifts simultaneously. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 4419–4430, 2024

    work page 2024

  5. [5]

    Dynamic environment responsive online meta-learning with fairness awareness

    Chen Zhao, Feng Mi, Xintao Wu, Kai Jiang, Latifur Khan, and Feng Chen. Dynamic environment responsive online meta-learning with fairness awareness. ACM Transactions on Knowledge Discovery from Data, 18(6):1–23, 2024

    work page 2024

  6. [6]

    Learning fair invariant representations under covariate and correlation shifts simultaneously

    Dong Li, Chen Zhao, Minglai Shao, and Wenjun Wang. Learning fair invariant representations under covariate and correlation shifts simultaneously. In Proceedings of the 33rd ACM International Conference on Information and Knowledge Management, pages 1174–1183, 2024

    work page 2024

  7. [7]

    Towards counterfactual fairness-aware domain generalization in changing environments

    Yujie Lin, Chen Zhao, Minglai Shao, Baoluo Meng, Xujiang Zhao, and Haifeng Chen. Towards counterfactual fairness-aware domain generalization in changing environments. IJCAI, 2024

    work page 2024

  8. [8]

    Adaptation speed analysis for fairness- aware causal models

    Yujie Lin, Chen Zhao, Minglai Shao, Xujiang Zhao, and Haifeng Chen. Adaptation speed analysis for fairness- aware causal models. In Proceedings of the 32nd ACM International Conference on Information and Knowledge Management, pages 1421–1430, 2023

    work page 2023

Show all 66 references
  1. [9]

    Towards fair disentangled online learning for changing environments

    Chen Zhao, Feng Mi, Xintao Wu, Kai Jiang, Latifur Khan, Christan Grant, and Feng Chen. Towards fair disentangled online learning for changing environments. In Proceedings of the ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 2023

    2023

  2. [10]

    Adaptive fairness-aware online meta-learning for changing environments

    Chen Zhao, Feng Mi, Xintao Wu, Kai Jiang, Latifur Khan, and Feng Chen. Adaptive fairness-aware online meta-learning for changing environments. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 2565–2575, 2022

    2022

  3. [11]

    Fairness-aware online meta-learning

    Chen Zhao, Feng Chen, and Bhavani Thuraisingham. Fairness-aware online meta-learning. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, pages 2294–2304, 2021

    2021

  4. [12]

    Feed: Fairness-enhanced meta-learning for domain generalization

    Kai Jiang, Chen Zhao, Haoliang Wang, and Feng Chen. Feed: Fairness-enhanced meta-learning for domain generalization. arXiv preprint arXiv:2411.01316, 2024

    2024 arXiv

  5. [13]

    Domain generalization using causal matching

    Divyat Mahajan, Shruti Tople, and Amit Sharma. Domain generalization using causal matching. In International conference on machine learning, pages 7313–7324. PMLR, 2021

    2021

  6. [14]

    Causality inspired representation learning for domain generalization

    Fangrui Lv, Jian Liang, Shuang Li, Bin Zang, Chi Harold Liu, Ziteng Wang, and Di Liu. Causality inspired representation learning for domain generalization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 8046–8056, 2022

    2022

  7. [15]

    Flood: A flexible invariant learning framework for out-of-distribution generalization on graphs

    Yang Liu, Xiang Ao, Fuli Feng, Yunshan Ma, Kuan Li, Tat-Seng Chua, and Qing He. Flood: A flexible invariant learning framework for out-of-distribution generalization on graphs. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 1548–...

    2023

  8. [16]

    Handling distribution shifts on graphs: An invariance perspective

    Qitian Wu, Hengrui Zhang, Junchi Yan, and David Wipf. Handling distribution shifts on graphs: An invariance perspective. arXiv preprint arXiv:2202.02466, 2022

    2022 arXiv

  9. [17]

    Metropolis-hastings data augmentation for graph neural networks

    Hyeonjin Park, Seunghun Lee, Sihyeon Kim, Jinyoung Park, Jisu Jeong, Kyung-Min Kim, Jung-Woo Ha, and Hyunwoo J Kim. Metropolis-hastings data augmentation for graph neural networks. Advances in Neural Information Processing Systems, 34:19010–19020, 2021

    2021

  10. [18]

    Robust optimization as data augmentation for large-scale graphs

    Kezhi Kong, Guohao Li, Mucong Ding, Zuxuan Wu, Chen Zhu, Bernard Ghanem, Gavin Taylor, and Tom Goldstein. Robust optimization as data augmentation for large-scale graphs. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 60–69, 2022

    2022

  11. [19]

    Adversarial causal augmentation for graph covariate shift

    Yongduo Sui, Xiang Wang, Jiancan Wu, An Zhang, and Xiangnan He. Adversarial causal augmentation for graph covariate shift. arXiv preprint arXiv:2211.02843, 2022. 14 MLDGG: Meta-Learning for Domain Generalization on Graphs (Accepted in KDD 2025)

    2022 arXiv

  12. [20]

    Learning to learn with variational information bottleneck for domain generalization

    Yingjun Du, Jun Xu, Huan Xiong, Qiang Qiu, Xiantong Zhen, Cees GM Snoek, and Ling Shao. Learning to learn with variational information bottleneck for domain generalization. In Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Pa...

    2020

  13. [21]

    Causalvae: Disentangled representation learning via neural structural causal models

    Mengyue Yang, Furui Liu, Zhitang Chen, Xinwei Shen, Jianye Hao, and Jun Wang. Causalvae: Disentangled representation learning via neural structural causal models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 9593–9602, 2021

    2021

  14. [22]

    Causal attention for interpretable and generalizable graph classification

    Yongduo Sui, Xiang Wang, Jiancan Wu, Min Lin, Xiangnan He, and Tat-Seng Chua. Causal attention for interpretable and generalizable graph classification. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, pages 1696–1705, 2022

    2022

  15. [23]

    Learning invariant graph representations for out-of- distribution generalization

    Haoyang Li, Ziwei Zhang, Xin Wang, and Wenwu Zhu. Learning invariant graph representations for out-of- distribution generalization. Advances in Neural Information Processing Systems, 35:11828–11841, 2022

    2022

  16. [24]

    Energy-based out-of-distribution detection for graph neural networks

    Qitian Wu, Yiting Chen, Chenxiao Yang, and Junchi Yan. Energy-based out-of-distribution detection for graph neural networks. In International Conference on Learning Representations (ICLR), 2023

    2023

  17. [25]

    Multi-scale attributed node embedding

    Benedek Rozemberczki, Carl Allen, and Rik Sarkar. Multi-scale attributed node embedding. Journal of Complex Networks, 9(2):cnab014, 2021

    2021

  18. [26]

    Social structure of facebook networks

    Amanda L Traud, Peter J Mucha, and Mason A Porter. Social structure of facebook networks. Physica A: Statistical Mechanics and its Applications, 391(16):4165–4180, 2012

    2012

  19. [27]

    Geom-gcn: Geometric graph convolutional networks

    Hongbin Pei, Bingzhe Wei, Kevin Chen-Chuan Chang, Yu Lei, and Bo Yang. Geom-gcn: Geometric graph convolutional networks. arXiv preprint arXiv:2002.05287, 2020

    2002 arXiv

  20. [28]

    Graphglow: Universal and generalizable structure learning for graph neural networks

    Wentao Zhao, Qitian Wu, Chenxiao Yang, and Junchi Yan. Graphglow: Universal and generalizable structure learning for graph neural networks. arXiv preprint arXiv:2306.11264, 2023

    2023 arXiv

  21. [29]

    Model-based domain generalization

    Alexander Robey, George J Pappas, and Hamed Hassani. Model-based domain generalization. Advances in Neural Information Processing Systems, 34:20210–20229, 2021

    2021

  22. [30]

    Learn to expect the unexpected: Probably approximately correct domain generalization

    Vikas Garg, Adam Tauman Kalai, Katrina Ligett, and Steven Wu. Learn to expect the unexpected: Probably approximately correct domain generalization. In International Conference on Artificial Intelligence and Statistics, pages 3574–3582. PMLR, 2021

    2021

  23. [31]

    Domain generalization: A survey

    Kaiyang Zhou, Ziwei Liu, Yu Qiao, Tao Xiang, and Chen Change Loy. Domain generalization: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022

    2022

  24. [32]

    Robust optimization over multiple domains

    Qi Qian, Shenghuo Zhu, Jiasheng Tang, Rong Jin, Baigui Sun, and Hao Li. Robust optimization over multiple domains. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 33, pages 4739–4746, 2019

    2019

  25. [33]

    Distributionally robust neural networks for group shifts: On the importance of regularization for worst-case generalization

    Shiori Sagawa, Pang Wei Koh, Tatsunori B Hashimoto, and Percy Liang. Distributionally robust neural networks for group shifts: On the importance of regularization for worst-case generalization. arXiv preprint arXiv:1911.08731, 2019

    1911 arXiv

  26. [34]

    Out-of-distribution generalization via risk extrapolation (rex)

    David Krueger, Ethan Caballero, Joern-Henrik Jacobsen, Amy Zhang, Jonathan Binas, Dinghuai Zhang, Remi Le Priol, and Aaron Courville. Out-of-distribution generalization via risk extrapolation (rex). In International Conference on Machine Learning, pages 5815–5826. PMLR, 2021

    2021

  27. [35]

    Learning to generalize: Meta-learning for domain generalization

    Da Li, Yongxin Yang, Yi-Zhe Song, and Timothy Hospedales. Learning to generalize: Meta-learning for domain generalization. In Proceedings of the AAAI conference on artificial intelligence, volume 32, 2018

    2018

  28. [36]

    Discriminative adversarial domain generalization with meta-learning based cross-domain validation

    Keyu Chen, Di Zhuang, and J Morris Chang. Discriminative adversarial domain generalization with meta-learning based cross-domain validation. Neurocomputing, 467:418–426, 2022

    2022

  29. [37]

    Cross-domain few-shot graph classification

    Kaveh Hassani. Cross-domain few-shot graph classification. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, pages 6856–6864, 2022

    2022

  30. [38]

    Adapting distilled knowledge for few-shot relation reasoning over knowledge graphs

    Yiming Zhang, Yiyue Qian, Yanfang Ye, and Chuxu Zhang. Adapting distilled knowledge for few-shot relation reasoning over knowledge graphs. In Proceedings of the 2022 SIAM International Conference on Data Mining (SDM), pages 666–674. SIAM, 2022

    2022

  31. [39]

    Model-agnostic meta-learning for fast adaptation of deep networks

    Chelsea Finn, Pieter Abbeel, and Sergey Levine. Model-agnostic meta-learning for fast adaptation of deep networks. In International conference on machine learning, pages 1126–1135. PMLR, 2017

    2017

  32. [40]

    A perspective view and survey of meta-learning

    Ricardo Vilalta and Youssef Drissi. A perspective view and survey of meta-learning. Artificial intelligence review, 18:77–95, 2002. 15 MLDGG: Meta-Learning for Domain Generalization on Graphs (Accepted in KDD 2025)

    2002

  33. [41]

    Episodic training for domain generalization

    Da Li, Jianshu Zhang, Yongxin Yang, Cong Liu, Yi-Zhe Song, and Timothy M Hospedales. Episodic training for domain generalization. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 1446–1455, 2019

    2019

  34. [42]

    Metareg: Towards domain generalization using meta-regularization

    Yogesh Balaji, Swami Sankaranarayanan, and Rama Chellappa. Metareg: Towards domain generalization using meta-regularization. Advances in neural information processing systems, 31, 2018

    2018

  35. [43]

    Domain generalization via model-agnostic learning of semantic features

    Qi Dou, Daniel Coelho de Castro, Konstantinos Kamnitsas, and Ben Glocker. Domain generalization via model-agnostic learning of semantic features. Advances in neural information processing systems, 32, 2019

    2019

  36. [44]

    Graph meta learning via local subgraphs

    Kexin Huang and Marinka Zitnik. Graph meta learning via local subgraphs. Advances in neural information processing systems, 33:5862–5874, 2020

    2020

  37. [45]

    Meta-gnn: On few- shot node classification in graph meta-learning

    Fan Zhou, Chengtai Cao, Kunpeng Zhang, Goce Trajcevski, Ting Zhong, and Ji Geng. Meta-gnn: On few- shot node classification in graph meta-learning. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management, pages 2357–2360, 2019

    2019

  38. [46]

    Graph few-shot learning via knowledge transfer

    Huaxiu Yao, Chuxu Zhang, Ying Wei, Meng Jiang, Suhang Wang, Junzhou Huang, Nitesh Chawla, and Zhenhui Li. Graph few-shot learning via knowledge transfer. In Proceedings of the AAAI conference on artificial intelligence, volume 34, pages 6656–6663, 2020

    2020

  39. [47]

    Graph few-shot learning with attribute matching

    Ning Wang, Minnan Luo, Kaize Ding, Lingling Zhang, Jundong Li, and Qinghua Zheng. Graph few-shot learning with attribute matching. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management, pages 1545–1554, 2020

    2020

  40. [48]

    Adaptive-step graph meta-learner for few-shot graph classification

    Ning Ma, Jiajun Bu, Jieyu Yang, Zhen Zhang, Chengwei Yao, Zhi Yu, Sheng Zhou, and Xifeng Yan. Adaptive-step graph meta-learner for few-shot graph classification. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management, pages 1055–1064, 2020

    2020

  41. [49]

    Few-shot graph learning for molecular property prediction

    Zhichun Guo, Chuxu Zhang, Wenhao Yu, John Herr, Olaf Wiest, Meng Jiang, and Nitesh V Chawla. Few-shot graph learning for molecular property prediction. In Proceedings of the web conference 2021, pages 2559–2567, 2021

    2021

  42. [50]

    Learning to compare: Relation network for few-shot learning

    Flood Sung, Yongxin Yang, Li Zhang, Tao Xiang, Philip HS Torr, and Timothy M Hospedales. Learning to compare: Relation network for few-shot learning. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1199–1208, 2018

    2018

  43. [51]

    Graph few-shot class-incremental learning

    Zhen Tan, Kaize Ding, Ruocheng Guo, and Huan Liu. Graph few-shot class-incremental learning. In Proceedings of the Fifteenth ACM International Conference on Web Search and Data Mining, pages 987–996, 2022

    2022

  44. [52]

    Probabilistic model-agnostic meta-learning

    Chelsea Finn, Kelvin Xu, and Sergey Levine. Probabilistic model-agnostic meta-learning. Advances in neural information processing systems, 31, 2018

    2018

  45. [53]

    Generalizing from a few examples: A survey on few-shot learning

    Yaqing Wang, Quanming Yao, James T Kwok, and Lionel M Ni. Generalizing from a few examples: A survey on few-shot learning. ACM computing surveys (csur), 53(3):1–34, 2020

    2020

  46. [54]

    Evolutionary principles in self-referential learning, or on learning how to learn: the meta-meta-

    Jürgen Schmidhuber. Evolutionary principles in self-referential learning, or on learning how to learn: the meta-meta-... hook. PhD thesis, Technische Universität München, 1987

    1987

  47. [55]

    Learning to learn single domain generalization

    Fengchun Qiao, Long Zhao, and Xi Peng. Learning to learn single domain generalization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12556–12565, 2020

    2020

  48. [56]

    How to train your maml

    Antreas Antoniou, Harri Edwards, and Amos Storkey. How to train your maml. In Seventh International Conference on Learning Representations, 2019

    2019

  49. [57]

    Multi-domain generalized graph meta learning

    Mingkai Lin, Wenzhong Li, Ding Li, Yizhou Chen, Guohao Li, and Sanglu Lu. Multi-domain generalized graph meta learning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 37, pages 4479–4487, 2023

    2023

  50. [58]

    Sample efficient reinforcement learning with reinforce

    Junzi Zhang, Jongho Kim, Brendan O’Donoghue, and Stephen Boyd. Sample efficient reinforcement learning with reinforce. In Proceedings of the AAAI conference on artificial intelligence, volume 35, pages 10887–10895, 2021

    2021

  51. [59]

    Auto-encoding variational bayes

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

    2013 arXiv

  52. [60]

    An introduction to variational methods for graphical models

    Michael I Jordan, Zoubin Ghahramani, Tommi S Jaakkola, and Lawrence K Saul. An introduction to variational methods for graphical models. Machine learning, 37:183–233, 1999

    1999

  53. [61]

    Analysis of the cholesky decomposition of a semi-definite matrix

    Nicholas J Higham. Analysis of the cholesky decomposition of a semi-definite matrix. 1990

    1990

  54. [62]

    A new metric for probability distributions

    Dominik Maria Endres and Johannes E Schindelin. A new metric for probability distributions. IEEE Transactions on Information theory, 49(7):1858–1860, 2003. 16 MLDGG: Meta-Learning for Domain Generalization on Graphs (Accepted in KDD 2025)

    2003

  55. [63]

    Shift-robust gnns: Overcoming the limitations of localized graph training data

    Qi Zhu, Natalia Ponomareva, Jiawei Han, and Bryan Perozzi. Shift-robust gnns: Overcoming the limitations of localized graph training data. Advances in Neural Information Processing Systems, 34:27965–27977, 2021

    2021

  56. [64]

    Mixup for node and graph classification

    Yiwei Wang, Wei Wang, Yuxuan Liang, Yujun Cai, and Bryan Hooi. Mixup for node and graph classification. In Proceedings of the Web Conference 2021, pages 3663–3674, 2021

    2021

  57. [65]

    mixup: Beyond empirical risk minimization

    Hongyi Zhang, Moustapha Cisse, Yann N Dauphin, and David Lopez-Paz. mixup: Beyond empirical risk minimization. arXiv preprint arXiv:1710.09412, 2017

    2017 arXiv

  58. [66]

    Semi-supervised classification with graph convolutional networks

    Thomas N Kipf and Max Welling. Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907, 2016. 17 MLDGG: Meta-Learning for Domain Generalization on Graphs (Accepted in KDD 2025) A Notations For clear interpretation, we list the notation...

    2016 arXiv

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.