Before It's Too Late: A State Space Model for the Early Prediction of Misinformation and Disinformation Engagement

Emily Booth; Francesco Bailo; Julian Droogan; Lin Tian; Marian-Andrei Rizoiu

REVIEW 3 major objections 5 minor 1 cited by

Before It's Too Late: A State Space Model for the Early Prediction of Misinformation and Disinformation Engagement

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 IC-Mamba forecasts engagement with misinformation within the first 15–30 minutes after posting, and carries the forecast to opinion-level curves 28 days ahead.

desk verdict A real Mamba adaptation for interval-censored engagement, but the pretraining protocol is underspecified enough that the reported 4.72% gain is not trustworthy until the split is confirmed. read the letter →

arxiv 2502.04655 v1 pith:ORWVIWYC submitted 2025-02-07 cs.CL

Lin Tian , Emily Booth , Francesco Bailo , Julian Droogan , Marian-Andrei Rizoiu This is my paper · ORCID

classification cs.CL

keywords statespacemodelinterval-censoreddataearlypredictionengagementforecastingmisinformationdisinformationopinionclassificationMamba

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 introduces IC-Mamba, a state space model—that is, a sequence model that advances a hidden state over time—that forecasts social-media engagement by treating engagement counts as interval-censored: each observation is a cumulative count at a discrete time, hiding the exact instants of likes, shares, comments, and emoji reactions. Its stated goal is early warning: within the first 15–30 minutes after posting it reaches RMSE between 0.118 and 0.143, and with 3–10 day observation windows it predicts opinion-level engagement up to 28 days ahead. Across four misinformation themes and a disinformation network, the authors report a 4.72% improvement over the best baselines on engagement metrics and F1 scores of 0.508 to 0.751 for classifying posts into opinion or narrative classes. If these results hold, platforms and researchers would have extra lead time to intervene before harmful content goes viral.

What carries the argument

The load-bearing component is the interval-aware state representation $v_j = [\Delta^-_j; \log(1+e_j); \Delta^+_j; \log(1+\hat{e}_{j+1})]$, which encodes how long since the last observation, the observed engagement in the current interval, the forward interval length, and the predicted next engagement. This vector feeds a selective state space model with time-dependent transitions $A_t(\Delta t) = \exp(\Delta t \tilde{A}_t)$, so the hidden state advances by the actual elapsed time rather than by a fixed tick. Time-aware positional embeddings combine relative and absolute encodings, modulated by engagement level, and a two-tier architecture carries predictions from individual posts to aggregate opinion curves. Pretraining on 1.78 million posts with an engagement-prediction loss plus a temporal-coherence loss supplies the initialization that makes fine-tuning on smaller labeled sets feasible.

What would settle it

Reproduce IC-Mamba with pre-training restricted to the earliest 70% of chronologically ordered posts, the same training period used for fine-tuning, and check that no validation- or test-period post appears in the pre-training corpus; if the 4.72% advantage over the best baseline shrinks or disappears, the central claim fails. A simpler audit is to compare the timestamps of the pretraining corpus against the test period.

Watch

Extended reading notes

Core claim

The central claim is that interval-censored engagement histories can be embedded directly into a selective state space model so that the model learns fine-grained temporal dynamics of engagement growth, and that this design forecasts engagement better than transformer, Hawkes, and plain Mamba baselines. IC-Mamba predicts the future trajectory of likes, shares, comments, and emojis for each post, sums those trajectories to opinion-level curves, and classifies posts into predefined opinions. On the paper's temporal holdout, it reports the best RMSE, MAPE, and $R^2$ on every dataset, with the largest qualitative advantage in the first hour after posting.

Load-bearing premise

The reported advantage depends on the large set of posts used to pre-train the model containing none of the posts used later to validate or test it; if test-period posts were included in pre-training, the gains could come from data leakage rather than from the model's design.

Editorial extensions

If this is right

Early reach assessment: from 15 minutes of observed engagement, IC-Mamba predicts the trajectory with RMSE 0.118–0.143, so a post's likely reach can be judged before it peaks.
Opinion-level early warning: with 3-, 7-, or 10-day observation windows, the two-tier model forecasts an opinion's total engagement up to 28 days ahead, and the reported confidence intervals tighten as more data streams in.
Better narrative classification than text-only models: on the disinformation network dataset, IC-Mamba's F1 of 0.508 substantially exceeds the reported text-only baseline of 0.107, indicating temporal and user features carry much of the signal in coordinated campaigns.
Ablations locate the signal: removing text hurts most on organic SocialSense content, removing temporal features hurts most on the coordinated DiN dataset, and removing user features hurts uniformly.
A reusable pretraining recipe: pretraining on 1.78 million posts lets the model fine-tune on smaller labeled misinformation sets, which matters when expert-labeled campaigns are scarce.

Reading between the lines

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

Editorial extension: the interval-aware representation is not tied to Facebook; the same construction should transfer to any platform that exposes only binned counts, such as view clocks or reshare totals.
Editorial extension: if the temporal-coherence loss genuinely enforces consistent hidden-state transitions across irregular intervals, IC-Mamba could also serve as a generative simulator of engagement trajectories for counterfactual intervention experiments.
Editorial extension: the paper's power-law analysis hints at a deployable discriminator: naturally diffusing content plateaus near human-scale engagement limits while coordinated content reaches far higher counts, so the plateau itself could flag suspected artificial amplification.
Editorial extension: because IC-Mamba's reported advantage over baselines grows with the length of the observation window, a head-to-head test on longer horizons, such as 90 days, would show whether the gap keeps widening.

Share X Bluesky LinkedIn Reddit HN

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

Colleague,

The paper is a genuine extension of Mamba to interval-censored engagement data, with time-aware embeddings and a two-tier post/opinion architecture, and it ships code and a dashboard. That's real, and the early-prediction focus (15–30 minutes) is a useful selling point. But the evaluation has a potentially critical ambiguity: Section 3.5 defines the pretraining corpus as all 1.78M posts 'collected from the two datasets,' while the chronological 70/15/15 split is described only for fine-tuning. If pretraining includes test-period posts, the model has seen the trajectories it is later asked to forecast, and the consistent 4–5% RMSE gains in Table 3 become a leakage artifact. The paper's statement that 'no future information leaks into training' does not explicitly cover the pretraining stage. This is the load-bearing question, and it needs an explicit answer from the authors.

What the paper does well: the architecture is sound and the ablations (removing text, user, time) show consistent contributions across both SocialSense and DiN. The interval-aware state representation and the temporal coherence regularizer are sensible adaptations of selective SSMs to irregularly sampled data. The opinion-level forecasting at 28 days with 3–10 day windows is a useful demonstration.

Soft spots beyond the split question: no standard deviations are reported anywhere despite Appendix B claiming ten runs; the small gaps (0.118 vs 0.124) may be within noise. The engagement normalization is never stated, so RMSE units are ambiguous. The exclusions (posts with fewer than four intervals, opinions with fewer than 5,000 posts) are post hoc and should be justified.

If the pretraining split is confirmed, this is a solid applied contribution. As written, the headline claim is not supported by the evidence in the text. That is a fixable problem, not a fatal one. The paper deserves a serious referee; the code is available, the datasets are real, and the questions are answerable. My recommendation: send it to review with a request that the authors clarify or redo the pretraining split, report variance, and define normalization.

Referee Report

3 major / 5 minor

Summary. The paper introduces IC-Mamba, a state space model that extends Mamba with time-aware positional embeddings, interval-aware state representations, and a temporal-coherence regularizer, and applies it to forecasting Facebook engagement (likes, shares, comments, emojis) for misinformation and disinformation content. The authors report post-level engagement prediction improvements over eight baselines (RMSE 0.118–0.143 across datasets in Table 3), early prediction experiments from 15 minutes to 6 hours, opinion-level classification (F1 0.508–0.751), and illustrative 28-day dynamic forecasts for an opinion-level trajectory. The central claim is a 4.72% improvement over state-of-the-art engagement forecasting.

Significance. If the evaluation is valid, the paper makes a practical contribution to early detection of misinformation and disinformation by integrating interval-censored engagement data with a modern state space architecture. Strengths include the use of two real-world Facebook datasets, a temporal holdout protocol, ablations, and a released code repository and dashboard. However, the current manuscript leaves a critical ambiguity about whether pretraining respects the temporal split, and it overstates the early-window results; these issues must be resolved before the claimed improvements can be trusted.

major comments (3)

[Section 3.5 / Section 4.2] The pretraining corpus D is defined in Section 3.5 as 1.78 million posts 'collected from the two datasets' with their complete engagement timelines, which per Table 2 is all posts in SocialSense and DiN. The temporal holdout in Section 4.2 (70/15/15 chronological split) is described only for the training/validation/test phases after pretraining. If Algorithm 1 is run on all posts including the test-period 15%, the model has already seen the engagement trajectories it is later asked to forecast, and the reported gains in Table 3 (e.g., Bushfire RMSE 0.118 vs. Mamba 0.124) could be leakage artifacts. The manuscript must state explicitly whether pretraining is restricted to the earliest 70% of posts; if it is not, the experiments must be rerun with split-aware pretraining before the 4.72% improvement claim can be accepted.
[Abstract / Section 4.3 / Section 4.4] The repeated claim that IC-Mamba excels at predicting engagement 'within the crucial first 15-30 minutes of posting (RMSE 0.118-0.143)' is not supported by the reported numbers. Table 3 and Section 4.3 use a 6-hour observation window to forecast cumulative engagement; the 0.118–0.143 range is exactly the Table 3 RMSE for that setting. Figure 5a shows that RMSE at 15 minutes is considerably higher, and Table 8 reports early-stage (next interval ≤1 hour) RMSE values of 0.169–0.235. The abstract and RQ2 should be corrected to state the actual observation window and horizon associated with the 0.118–0.143 RMSE, or the experiments should be rerun to support the early-window claim.
[Section 4.5 / RQ3] The 28-day opinion-level forecasting results are presented only as illustrative plots in Fig. 5(b)(c), without quantitative error metrics, comparison baselines, or ablations. The abstract claims the model 'successfully forecasting opinion-level engagement up to 28 days ahead using observation windows of 3-10 days,' but no evaluation table or metric is provided for this scenario. The manuscript should either add quantitative results (e.g., RMSE/MAPE over the forecast horizon, with the same baselines as Table 3) or explicitly soften the claim to reflect that the 28-day results are demonstrative only.

minor comments (5)

[Appendix B.1 / Table 3] Appendix B.1 states that all reported figures are averaged over ten runs with different random seeds, but Table 3 reports no variances or significance tests; adding standard deviations or confidence intervals would strengthen the claim that IC-Mamba outperforms every baseline on every metric.
[Table 5] Table 5 reports hyperparameter ranges only; the final selected values and the selection procedure (grid search, Bayesian optimization, etc.) should be reported to make the experiments reproducible.
[Section 4.5 / Figure 5] The observation windows are described inconsistently: Section 4.5 and Figure 5 use 7-day and 10-day windows, while the abstract says '3-10 days.' Clarify which windows are actually used in the 28-day forecasts.
[Section 4.1] The sentence 'Fig. 4(c) offers examines comment distributions' contains a typo and should be revised to 'examines' or 'offers an examination of.'
[Figure 5] The 95% confidence intervals are described as 'obtained from all previous prediction for this time'; the computation should be clarified, including whether they come from the ten random seeds, Monte Carlo sampling, or empirical quantiles.

Circularity Check

2 steps flagged · score 7.0 of 10

IC-Mamba's 'prediction' is self-referential: the target engagement vector ê_{j+1} is an explicit component of the input v_j used to predict it, so the reported RMSE gains do not measure out-of-sample forecasting.

self definitional [Section 3.4, Interval-aware State Representation; Section 3.5, Eq. (1)]
"For each observation time t_j in the engagement history H_τobs(p), we construct an interval-aware vector v_j ∈ R^{4d}: v_j =[Δt^-_j ; log(1+e_j); Δt^+_j ; log(1+ ê_{j+1})] ... and ê_{j+1} is the predicted next engagement vector. Lpred = 1/|P| ∑_{p∈P} ∑_{j=0}^{m−1} ∥ê_{j+1}−e_{j+1}∥^2"

Equation (1) trains the model to produce ê_{j+1}, but the same vector ê_{j+1} is already a component of the interval-aware input v_j at the same time index j. The SSM output at position j is a function of inputs up to j, so minimizing Lpred can be achieved by copying the provided ê_{j+1} component; the loss measures reproduction of an input feature, not forecasting. The prediction-time formula v_k = [τ_k−t_j; log(1+e_j); τ_{k+1}−τ_k; log(1+ê_k)] is likewise self-referential unless an initialization or fixed-point procedure is specified, which the paper does not provide.
fitted input called prediction [Section 3.5 (pretraining corpus D) vs Section 4.2 (temporal holdout)]
"We introduce D = {(p_i, H_i, x_i, u_i)}^M_{i=1}, a pretraining dataset comprising 1.78 million posts and their associated social engagement timelines ... collected from the two datasets SocialSense and DiN. We chronologically order all posts and use the earliest 70% for training, the next 15% for validation, and the most recent 15% for testing."

Table 2's post counts sum to approximately 1.78 million, so D is all posts from both datasets, including the most recent 15% held out as the test set in Section 4.2. The paper does not state that pretraining is restricted to the earliest 70%; Algorithm 1 iterates over batches B⊂D. If pretraining sees test-period engagement timelines, the model's test RMSE is a fit to data already used in training, not an out-of-sample prediction. This turns the reported 4.72% improvement into a potential leakage artifact.

full rationale

The central empirical claim is undermined by a direct self-definitional construction in Section 3.4: the quantity being predicted appears as an input. This is not a matter of author intent or a self-citation; it is visible in the equations. Even if the pretraining-corpus ambiguity is resolved in the authors' favor, the v_j/ê_{j+1} overlap makes the training objective Eq. (1) unable to certify forecasting ability. The temporal coherence loss is a consistency regularizer and is not separately circular. The pretraining D definition additionally creates a leakage risk because the reported split is only described for fine-tuning. These are the only significant circular steps; no load-bearing self-citation or imported uniqueness theorem is present, and the comparison against external baselines would be meaningful only if the target-in-input leak were corrected.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim is an empirical performance claim from a trained model. It rests on several hand-chosen constants (sigma, lambda, granularity, observation window, thresholds) and on unstated evaluation normalization. The most consequential assumption is also procedural: pretraining must not see the test period, which the paper does not explicitly guarantee.

free parameters (5)

RTE scale sigma = learned; exact value not reported
RTE(t,t_ref)=sin((t-t_ref)/sigma) where sigma is a learnable parameter controlling the temporal scale; no ablation on sigma is given.
Temporal coherence loss weight lambda = lambda ranges in (0.1,0.9) or (0.4,0.6) (Table 5)
Balances engagement prediction loss and temporal-coherence loss in pretraining; chosen by validation and directly shapes the learned state transitions.
Time granularity (prediction step) = 5 minutes or 1 hour
Choice of fixed interval for predictions; all RMSE numbers depend on this discretization.
Observation window tau_obs = 6 hours for post-level; 3, 7, or 10 days for opinion-level
Observation window length determines the amount of history used and is central to the early prediction claims.
Minimum engagement intervals and opinion size thresholds = 4 intervals; 5,000 posts
Post hoc filtering of evaluation data; affects which posts and opinions are scored.

assumptions (4)

ad hoc to paper Engagement counts are transformed (e.g., log(1+count)) before prediction and before computing RMSE, MAPE, and R2, but the exact evaluation normalization is not stated.
The paper uses log(1+e) in interval-aware vectors and embeddings, but Section 4.3 does not specify the scale on which errors are computed; with counts reaching 10^6, unnormalized RMSE would not be around 0.1.
domain assumption CrowdTangle interval observations are an unbiased representation of actual engagement.
The model treats missing times as censored intervals and assumes the observed counts are the true cumulative counts; no validation against platform-side logs is provided (Section 4.1).
domain assumption Expert-assigned opinion and narrative labels are reliable ground truth.
SocialSense and DiN labels come from expert content analysis, but no inter-annotator agreement or label noise analysis is reported (Section 4.1).
standard math The matrix exponential transition exp(Delta t * A_t) is an appropriate generative model for engagement dynamics.
This is the standard continuous-time SSM discretization used in Section 3.4; it is assumed rather than derived from engagement data.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Before It's Too Late: A State Space Model for the Early Prediction of Misinformation and Disinformation Engagement." pith.science (2026). https://pith.science/paper/ORWVIWYC

@misc{pith2026250204655,
  author       = {Pith},
  title        = {Pith review of: Before It's Too Late: A State Space Model for the Early Prediction of Misinformation and Disinformation Engagement},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ORWVIWYC}},
  note         = {Machine review of arXiv:2502.04655}
}

read the original abstract

In today's digital age, conspiracies and information campaigns can emerge rapidly and erode social and democratic cohesion. While recent deep learning approaches have made progress in modeling engagement through language and propagation models, they struggle with irregularly sampled data and early trajectory assessment. We present IC-Mamba, a novel state space model that forecasts social media engagement by modeling interval-censored data with integrated temporal embeddings. Our model excels at predicting engagement patterns within the crucial first 15-30 minutes of posting (RMSE 0.118-0.143), enabling rapid assessment of content reach. By incorporating interval-censored modeling into the state space framework, IC-Mamba captures fine-grained temporal dynamics of engagement growth, achieving a 4.72% improvement over state-of-the-art across multiple engagement metrics (likes, shares, comments, and emojis). Our experiments demonstrate IC-Mamba's effectiveness in forecasting both post-level dynamics and broader narrative patterns (F1 0.508-0.751 for narrative-level predictions). The model maintains strong predictive performance across extended time horizons, successfully forecasting opinion-level engagement up to 28 days ahead using observation windows of 3-10 days. These capabilities enable earlier identification of potentially problematic content, providing crucial lead time for designing and implementing countermeasures. Code is available at: https://github.com/ltian678/ic-mamba. An interactive dashboard demonstrating our results is available at: https://ic-mamba.behavioral-ds.science.

Figures

Figures reproduced from arXiv: 2502.04655 by the authors.

**Figure 2.** Overview of the IC-Mamba Architecture for social media engagement prediction. (left panel) The model first takes [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Two-Tier IC-Mamba Architecture. The bottomtier model (IC-Mamba1) learns post-level representations from historical (𝐻), content (𝑥), and user (𝑢) features, while the top-tier model (IC-Mamba2) captures temporal dependencies across intervals 𝛿𝑡 to jointly predict individual post virality and aggregate narrative engagement dynamics [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Engagement distribution patterns across social media content. (a) Log-scale ECCDF of engagement metrics for the [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

**Figure 5.** Figure 5: Comparative analysis of early prediction performance and dynamic forecasting. (a) Performance comparison on [PITH_FULL_IMAGE:figures/full_fig_p008_5.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

Estimating Online Influence Needs Causal Modeling! Counterfactual Analysis of Social Media Engagement
cs.CL 2025-05 reject novelty 5.0 of 10

The paper proposes causal-Mamba, a joint treatment-outcome deep sequential model, to estimate average treatment effects of Google Trends-like external signals on social media engagement and to rank influential users.

Reference graph

Works this paper leans on

52 extracted references · 41 canonical work pages · cited by 1 Pith paper

[1]

[n. d.]. CrowdTangle. https://www.crowdtangle.com/

work page
[2]

Atanas Atanasov, Gianmarco De Francisci Morales, and Preslav Nakov. 2019. Predicting the Role of Political Trolls in Social Media. In Proceedings of the 23rd Conference on Computational Natural Language Learning (CoNLL) . 1023–1034

work page 2019
[3]

Sidney Black, Stella Biderman, Eric Hallahan, Quentin Gregory Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Martin Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, and Samuel Weinbach. 2022. GPT-NeoX-20B: An Open-Source Autoregressive Language Model. In Challenges & ...

work page 2022
[4]

Pio Calderon, Rohit Ram, and Marian-Andrei Rizoiu. 2024. Opinion Market Model: Stemming Far-Right Opinion Spread Using Positive Interventions. In Proceedings of the International AAAI Conference on Web and Social Media , Vol. 18. 177–190

work page 2024
[5]

Pio Calderon and Marian-Andrei Rizoiu. 2024. What Drives Online Popularity: Author, Content or Sharers? Estimating Spread Dynamics with Bayesian Mixture Hawkes. 142–160. https://doi.org/10.1007/978-3-031-70362-1_9

work page doi:10.1007/978-3-031-70362-1_9 2024
[6]

Pio Calderon, Alexander Soen, and Marian-Andrei Rizoiu. 2025. Linking Across Data Granularity: Fitting Multivariate Hawkes Processes to Partially Interval- Censored Data. IEEE Transactions on Computational Social Systems 12 (2 2025), 25–37. Issue 1. https://doi.org/10.1109/TCSS.2024.3486117

arXiv 2025
[7]

Qi Cao, Huawei Shen, Keting Cen, Wentao Ouyang, and Xueqi Cheng. 2017. Deephawkes: Bridging the Gap between Prediction and Understanding of Infor- mation Cascades. In Proceedings of the 2017 ACM on Conference on Information and Knowledge Management. 1149–1158

work page 2017
[8]

Si-An Chen, Chun-Liang Li, Sercan O Arik, Nathanael Christian Yoder, and Tomas Pfister. 2023. TSMixer: An All-MLP Architecture for Time Series Forecast-ing. Transactions on Machine Learning Research (2023)

work page 2023

Show all 52 references

[9]

Aaron Clauset, Cosma Rohilla Shalizi, and Mark EJ Newman. 2009. Power-law distributions in empirical data. SIAM Review 51, 4 (2009), 661–703

2009
[10]

Tri Dao, Dan Fu, Stefano Ermon, Atri Rudra, and Christopher Ré. 2022. Flashatten- tion: Fast and Memory-Efficient Exact Attention with IO-Awareness. Advances in Neural Information Processing Systems 35 (2022), 16344–16359

2022
[11]

Tri Dao and Albert Gu. 2024. Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Duality. In Proceedings of the 41st International Conference on Machine Learning , Vol. 235. 10041–10071

2024
[12]

Keyan Ding, Ronggang Wang, and Shiqi Wang. 2019. Social Media Popularity Prediction: A Multiple Feature Fusion Approach with Deep Neural Networks. In Proceedings of the 27th ACM International Conference on Multimedia . 2682–2686

2019
[13]

Robin IM Dunbar. 1992. Neocortex size as a constraint on group size in primates. Journal of Human Evolution 22, 6 (1992), 469–493

1992
[14]

Stefan Elfwing, Eiji Uchibe, and Kenji Doya. 2018. Sigmoid-weighted linear units for neural network function approximation in reinforcement learning. Neural Networks 107 (2018), 3–11

2018
[15]

Albert Gu and Tri Dao. 2024. Mamba: Linear-Time Sequence Modeling with Selective State Spaces. In Conference on Language Modeling

2024
[16]

Albert Gu, Karan Goel, and Christopher Re. 2022. Efficiently Modeling Long Sequences with Structured State Spaces. In International Conference on Learning Representations

2022
[17]

Albert Gu, Isys Johnson, Karan Goel, Khaled Saab, Tri Dao, Atri Rudra, and Christopher Ré. 2021. Combining Recurrent, Convolutional, and Continuous- time Models with Linear State Space Layers. Advances in Neural Information Processing Systems 34 (2021), 572–585

2021
[18]

Ramin Hasani, Mathias Lechner, Alexander Amini, Daniela Rus, and Radu Grosu
[19]

Jane Im, Eshwar Chandrasekharan, Jackson Sargent, Paige Lighthammer, Taylor Denby, Ankit Bhargava, Libby Hemphill, David Jurgens, and Eric Gilbert. 2020. Still out there: Modeling and Identifying Russian Troll Accounts on Twitter. In 12th ACM Conference on Web Science . 1–10

2020
[20]

Quyu Kong, Emily Booth, Francesco Bailo, Amelia Johns, and Marian-Andrei Rizoiu. 2022. Slipping to the Extreme: A Mixed Method to Explain How Extreme Opinions Infiltrate Online Discussions. In Proceedings of the International AAAI Conference on Web and Social Media , Vol. 16. 524–535

2022
[21]

Quyu Kong, Pio Calderon, Rohit Ram, Olga Boichak, and Marian-Andrei Rizoiu
[22]

Quyu Kong, Rohit Ram, and Marian-Andrei Rizoiu. 2021. Evently: A Toolkit for Analyzing Online Users via Reshare Cascade Modeling. In Proceedings of the 14th ACM International Conference on Web Search and Data Mining (New York, NY, USA). ACM, 1097–1100. https://doi.org/10.1145/...

2021
[23]

Quyu Kong, Marian-Andrei Rizoiu, Siqi Wu, and Lexing Xie. 2018. Will This Video Go Viral? Explaining and Predicting the Popularity of Youtube Videos. In Companion of the The Web Conference 2018 on The Web Conference 2018 (WWW ’18) (Lyon, France). ACM Press, 175–178. https://do...

2018
[24]

Quyu Kong, Marian-Andrei Rizoiu, and Lexing Xie. 2020. Describing and Pre- dicting Online Items with Reshare Cascades via Dual Mixture Self-exciting Processes. In Proceedings of the 29th ACM International Conference on Infor- mation & Knowledge Management (New York, NY, USA). ...

2020
[25]

Quyu Kong, Marian-Andrei Rizoiu, and Lexing Xie. 2020. Modeling Informa- tion Cascades with Self-exciting Processes via Generalized Epidemic Models. In Proceedings of the 13th International Conference on Web Search and Data Mining (New York, NY, USA). ACM, 286–294. https://doi...

2020
[26]

David MJ Lazer, Matthew A Baum, Yochai Benkler, Adam J Berinsky, Kelly M Greenhill, Filippo Menczer, Miriam J Metzger, Brendan Nyhan, Gordon Penny- cook, David Rothschild, et al. 2018. The science of fake news. Science 359, 6380 (2018), 1094–1096

2018
[27]

Cheng Li, Jiaqi Ma, Xiaoxiao Guo, and Qiaozhu Mei. 2017. Deepcas: An End-to- end Predictor of Information Cascades. In Proceedings of the 26th International Conference on World Wide Web. 577–586

2017
[28]

Shiyang Li, Xiaoyong Jin, Yao Xuan, Xiyou Zhou, Wenhu Chen, Yu-Xiang Wang, and Xifeng Yan. 2019. Enhancing the Locality and Breaking the Memory Bottle- neck of Transformer on Time Series Forecasting. Advances in Neural Information Processing Systems 32 (2019)

2019
[29]

Xiaodong Lu, Shuo Ji, Le Yu, Leilei Sun, Bowen Du, and Tongyu Zhu. 2023. Continuous-time graph learning for cascade popularity prediction. InProceedings of the Thirty-Second International Joint Conference on Artificial Intelligence . 2224– 2232

2023
[30]

Luca Luceri, Valeria Pantè, Keith Burghardt, and Emilio Ferrara. 2024. Unmasking the Web of Deceit: Uncovering Coordinated Activity to Expose Information Operations on Twitter. In Proceedings of the ACM on Web Conference 2024 . 2530– 2541

2024
[31]

Mark EJ Newman. 2005. Power laws, Pareto distributions and Zipf’s law. Con- temporary Physics 46, 5 (2005), 323–351

2005
[32]

Jiezhong Qiu, Jian Tang, Hao Ma, Yuxiao Dong, Kuansan Wang, and Jie Tang
[33]

Syama Sundar Rangapuram, Matthias W Seeger, Jan Gasthaus, Lorenzo Stella, Yuyang Wang, and Tim Januschowski. 2018. Deep state space models for time series forecasting. Advances in neural information processing systems 31 (2018)

2018
[34]

Marian-Andrei Rizoiu, Alexander Soen, Shidi Li, Pio Calderon, Leanne J Dong, Aditya Krishna Menon, and Lexing Xie. 2022. Interval-censored Hawkes pro- cesses. Journal of Machine Learning Research 23, 338 (2022), 1–84

2022
[35]

Marian-Andrei Rizoiu, Lexing Xie, Scott Sanner, Manuel Cebrian, Honglin Yu, and Pascal Van Hentenryck. 2017. Expecting to be HIP: Hawkes Intensity Processes for Social Media Popularity. In Proceedings of the 26th International Conference on World Wide Web. 735–744

2017
[36]

Dietram A Scheufele and Nicole M Krause. 2019. Science audiences, misinfor- mation, and fake news. Proceedings of the National Academy of Sciences 116, 16 (2019), 7662–7669

2019
[37]

Lin Tian, Xiuzhen Zhang, and Jey Han Lau. 2021. Rumour detection via zero-shot cross-lingual transfer learning. In Machine Learning and Knowledge Discovery in Databases. Research Track: European Conference, ECML PKDD 2021, Bilbao, Spain, September 13–17, 2021, Proceedings, Par...

2021
[38]

Lin Tian, Xiuzhen Zhang, and Jey Han Lau. 2023. Metatroll: Few-shot Detection of State-Sponsored Trolls with Transformer Adapters. In Proceedings of the ACM Web Conference 2023. 1743–1753

2023
[39]

Lin Tian, Xiuzhen Jenny Zhang, and Jey Han Lau. 2022. DUCK: Rumour Detection on Social Media by Modelling User and Comment Propagation Networks. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language...

2022
[40]

Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention Is All You Need. Advances in Neural Information Processing Systems 30 (2017)

2017
[41]

Jia Wang, Vincent W Zheng, Zemin Liu, and Kevin Chen-Chuan Chang. 2017. Topological recurrent neural network for diffusion prediction. In 2017 IEEE inter- national conference on data mining (ICDM) . IEEE, 475–484

2017
[42]

Yongqing Wang, Huawei Shen, Shenghua Liu, Jinhua Gao, and Xueqi Cheng. 2017. Cascade dynamics modeling with attention-based recurrent neural network. In Proceedings of the 26th International Joint Conference on Artificial Intelligence . 2985–2991

2017
[43]

Haixu Wu, Jiehui Xu, Jianmin Wang, and Mingsheng Long. 2021. Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Fore- casting. Advances in Neural Information Processing Systems 34 (2021), 22419– 22430

2021
[44]

Siqi Wu, Marian-Andrei Rizoiu, and Lexing Xie. 2018. Beyond Views: Measuring and Predicting Engagement in Online Videos. In Proceedings of the International AAAI Conference on Web and Social Media , Vol. 12

2018
[45]

Xovee Xu, Fan Zhou, Kunpeng Zhang, Siyuan Liu, and Goce Trajcevski. 2021. Casflow: Exploring hierarchical structures and propagation uncertainty for cas- cade prediction. IEEE Transactions on Knowledge and Data Engineering 35, 4 WWW ’25, April 28-May 2, 2025, Sydney, NSW, Aust...

2021
[46]

Savvas Zannettou, Tristan Caulfield, Emiliano De Cristofaro, Michael Sirivianos, Gianluca Stringhini, and Jeremy Blackburn. 2019. Disinformation warfare: Un- derstanding state-sponsored trolls on Twitter and their influence on the web. In Companion Proceedings of the 2019 Worl...

2019
[47]

Rui Zhang, Christian Walder, Marian-Andrei Rizoiu, and Lexing Xie. 2019. Effi- cient Non-parametric Bayesian Hawkes Processes. In Proceedings of the Twenty- Eighth International Joint Conference on Artificial Intelligence (California). Inter- national Joint Conferences on Arti...

2019 doi
[48]

Haoyi Zhou, Shanghang Zhang, Jieqi Peng, Shuai Zhang, Jianxin Li, Hui Xiong, and Wancai Zhang. 2021. Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 35. 11106–11115

2021
[49]

Simiao Zuo, Haoming Jiang, Zichong Li, Tuo Zhao, and Hongyuan Zha. 2020. Transformer hawkes process. In International Conference on Machine Learning . PMLR, 11692–11702. A Ethics Considerations Our research exclusively uses publicly available Facebook data on CrowdTangle, adhe...

2020
[2018]

In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining

Deepinf: Social Influence Prediction with Deep Learning. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2110–2119
[2021]

In Proceedings of the AAAI Conference on Artificial Intelligence, Vol

Liquid Time-Constant Networks. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 35. 7657–7666
[2023]

In Proceedings of the ACM Web Conference

Interval-censored Transformer Hawkes: Detecting Information Operations using the Reaction of Social Systems. In Proceedings of the ACM Web Conference

Pith tools

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

[1] [1]

[n. d.]. CrowdTangle. https://www.crowdtangle.com/

work page

[2] [2]

Atanas Atanasov, Gianmarco De Francisci Morales, and Preslav Nakov. 2019. Predicting the Role of Political Trolls in Social Media. In Proceedings of the 23rd Conference on Computational Natural Language Learning (CoNLL) . 1023–1034

work page 2019

[3] [3]

Sidney Black, Stella Biderman, Eric Hallahan, Quentin Gregory Anthony, Leo Gao, Laurence Golding, Horace He, Connor Leahy, Kyle McDonell, Jason Phang, Michael Martin Pieler, USVSN Sai Prashanth, Shivanshu Purohit, Laria Reynolds, Jonathan Tow, Ben Wang, and Samuel Weinbach. 2022. GPT-NeoX-20B: An Open-Source Autoregressive Language Model. In Challenges & ...

work page 2022

[4] [4]

Pio Calderon, Rohit Ram, and Marian-Andrei Rizoiu. 2024. Opinion Market Model: Stemming Far-Right Opinion Spread Using Positive Interventions. In Proceedings of the International AAAI Conference on Web and Social Media , Vol. 18. 177–190

work page 2024

[5] [5]

Pio Calderon and Marian-Andrei Rizoiu. 2024. What Drives Online Popularity: Author, Content or Sharers? Estimating Spread Dynamics with Bayesian Mixture Hawkes. 142–160. https://doi.org/10.1007/978-3-031-70362-1_9

work page doi:10.1007/978-3-031-70362-1_9 2024

[6] [6]

Pio Calderon, Alexander Soen, and Marian-Andrei Rizoiu. 2025. Linking Across Data Granularity: Fitting Multivariate Hawkes Processes to Partially Interval- Censored Data. IEEE Transactions on Computational Social Systems 12 (2 2025), 25–37. Issue 1. https://doi.org/10.1109/TCSS.2024.3486117

arXiv 2025

[7] [7]

Qi Cao, Huawei Shen, Keting Cen, Wentao Ouyang, and Xueqi Cheng. 2017. Deephawkes: Bridging the Gap between Prediction and Understanding of Infor- mation Cascades. In Proceedings of the 2017 ACM on Conference on Information and Knowledge Management. 1149–1158

work page 2017

[8] [8]

Si-An Chen, Chun-Liang Li, Sercan O Arik, Nathanael Christian Yoder, and Tomas Pfister. 2023. TSMixer: An All-MLP Architecture for Time Series Forecast-ing. Transactions on Machine Learning Research (2023)

work page 2023