Do Temporal Link Predictors Need Learned Memory? A Smoothed-Count Baseline with a Handful of Parameters

Lisi Qarkaxhija; Ingo Scholtes

cs.LGSep 28, 2026

Do Temporal Link Predictors Need Learned Memory? A Smoothed-Count Baseline with a Handful of Parameters

Authors: Lisi Qarkaxhija, Ingo Scholtes

Organizations: Chair of Machine Learning for Complex Networks Center for Artificial Intelligence and Data Science (CAIDAS) Julius-Maximilians-Universität Würzburg, DE

Abstract

Many temporal link predictors summarize past interactions through learned node representations. We examine whether simple counts of recurring interaction patterns can provide competitive predictions without learning these representations. We propose a temporal link predictor based on statistical language modelling. It pools transition and co-occurrence counts across sources to predict links that a source has never formed. We smooth sparse estimates using destination frequencies or Kneser-Ney continuation counts. A shared log-linear rule combines these estimates with popularity, source history, and recency, without node embeddings. In our main evaluation, the model achieves the highest MRR among the compared methods on 7 out of 16 datasets from TGB and TGB-Seq. It also outperforms EdgeBank and Base3 on all 16 datasets and the heuristic family on 14. These gains extend to datasets designed to limit repeated edges. With only 9--13 learned parameters, our model provides a simple and competitive baseline for evaluating future neural temporal link predictors.

Figures & tables

Dataset	EdgeBank	Heuristic	Base3	marginal	continuation	MRR	Method
				Ours		Best neural
wiki	64.40	82.06	73.26	82.84 $\pm$ 0.04	82.72 $\pm$ 0.10	82.70	TPNet
uci	32.40	52.75	35.22	51.57 $\pm$ 0.32	51.25 $\pm$ 0.26	24.50	TNCN
enron	15.60	84.66	48.65	86.82 $\pm$ 0.01	86.83 $\pm$ 0.03	37.90	TNCN
subreddit	59.22	74.19	74.25	74.90 $\pm$ 0.06	74.79 $\pm$ 0.04	69.60	TNCN
lastfm	2.63	16.58	11.37	32.84 $\pm$ 0.18	32.87 $\pm$ 0.12	15.60	TNCN

Table 1: Test MRR (%) on TGB. Our entries are means and sample standard deviations over five runs. EdgeBank is the stronger of its two variants. Best neural gives the strongest reported neural benchmark score (Appendix A ). Bold marks the highest MRR per row.

Dataset	EdgeBank	Heuristic	Base3	marginal	continuation	MRR	Method
				Ours		Best neural
GoogleLocal	1.96	18.39	3.97	31.27 $\pm$ 0.02	31.42 $\pm$ 0.03	62.88	SGNN-HN
Yelp	9.77	30.81	11.99	53.21 $\pm$ 0.06	52.16 $\pm$ 0.09	72.69	CRAFT
Taobao	20.28	42.99	20.17	59.58 $\pm$ 0.03	58.90 $\pm$ 0.06	70.68	CRAFT
ML-20M	1.94	17.71	6.64	39.24 $\pm$ 0.15	39.00 $\pm$ 0.10	35.91	CRAFT
Flickr	1.96	38.38	34.81	62.68 $\pm$ 0.07	62.66 $\pm$ 0.08	62.34	CRAFT

Table 2: Test MRR (%) on TGB-Seq. Columns follow Table 1 . Count baselines use the TGB-Seq candidate sets.

Appendix figures & tables5 assets

Supplementary material from the paper’s appendix.

Appendix

Estimator	$a$	$b$	$c$	$d$
$\hat{P}(z)$	0.20	0.30	0.30	0.20
$\hat{P}_{\mathrm{c}}(z)$	0.14	0.43	0.29	0.14
$\hat{P}(z\mid u)$	0.20	0.38	0.38	0.03
$\hat{P}(z\mid v_{1}{=}c)$	0.10	0.65	0.15	0.10
$\hat{P}(z\mid v_{2}{=}b)$	0.07	0.10	0.77	0.07

Appendix

Table 3: Smoothed estimates on the worked stream of Figure 1 , computed by the model implementation (

\alpha=1

, counts undecayed). Conditional rows use marginal backoff. Bold marks the candidates each estimator favours.

Dataset	Marginal backoff	Continuation backoff
wiki	all three	all three
uci	all three	all three
enron	all three	all three
subreddit	reference	global recency
lastfm	all three	all three
review	co-occurrence	co-occurrence

Appendix

Table 4: Feature sets used in the main comparisons. Entries name additions to the reference configuration. These choices remain fixed across five runs.

Dataset	Updates	EdgeBank	Heuristic	Base3	Ours	MRR	Method
						Best neural
wiki	Event	64.40	82.06	73.26	82.84 $\pm$ 0.04	82.70	TPNet
	Batch 200	57.10	72.90	68.07	74.08 $\pm$ 0.09	82.70	TPNet
uci	Event	32.40	52.75	35.22	51.57 $\pm$ 0.32	24.50	TNCN
	Batch 200	22.16	37.49	28.22	38.55 $\pm$ 0.16	24.50	TNCN
enron	Event	15.60	84.66	48.65	86.83 $\pm$ 0.03	37.90	TNCN

Appendix

Table 5: Test MRR (%) on TGB under event-wise and batch-200 updates. EdgeBank, Heuristic, Base3, and Ours share the listed schedule and candidates. Our scores are means and sample standard deviations over five runs. Neural reference scores retain their original protocols and are repeated for comparison. Bold marks the highest MRR in each row.

Dataset	Updates	EdgeBank	Heuristic	Base3	Ours	MRR	Method
						Best neural
GoogleLocal	Event	1.96	18.39	3.97	31.42 $\pm$ 0.03	62.88	SGNN-HN
	Batch 200	1.96	18.39	3.73	29.00 $\pm$ 0.01	62.88	SGNN-HN
Yelp	Event	9.77	30.81	11.99	53.21 $\pm$ 0.06	72.69	CRAFT
	Batch 200	9.76	30.81	11.88	52.26 $\pm$ 0.05	72.69	CRAFT
Taobao	Event	20.28	42.99	20.17	59.58 $\pm$ 0.03	70.68	CRAFT

Appendix

Table 6: Test MRR (%) on TGB-Seq under event-wise and batch-200 updates. Columns and conventions follow Table 5 . Neural reference scores retain their original protocols.

TGB
Dataset	Event-wise	Strict timestamp	Change (points)
wiki	82.84 $\pm$ 0.04	82.84 $\pm$ 0.04	0.00 $\pm$ 0.00
uci	51.57 $\pm$ 0.32	51.43 $\pm$ 0.32	-0.14 $\pm$ 0.00
enron	86.83 $\pm$ 0.03	53.50 $\pm$ 0.18	-33.33 $\pm$ 0.16
subreddit	74.90 $\pm$ 0.06	74.89 $\pm$ 0.06	0.00 $\pm$ 0.00
lastfm	32.84 $\pm$ 0.18	32.78 $\pm$ 0.18	-0.06 $\pm$ 0.00

Appendix

Table 7: Strict-timestamp sensitivity of our model. Scores are test MRR (%), with means and sample standard deviations over five runs. Changes from event-wise evaluation use unrounded scores paired by seed. Configurations, weights, and candidates are fixed.

Explore similar work

EdgeReMIND: A Scalable, Top-Ranked Memorization Baseline for Temporal Multi-Relational Link Prediction
Sep 15, 2026Bryant PollardZero-Shot Link Prediction Memo
Clemson University
Where Does Neural Advantage Arise in Continuous-Time Dynamic Graph Prediction?
Aug 7, 2026Minwoo Yu, Young-guk HaZero-Shot Link Prediction Data-Driven Forecasting
Smart Computing Laboratory, Department of Computer Science & Engineering, Konkuk University, Seoul 05029, Republic of Korea
Estimation-Prediction Tradeoff in Causal Probabilistic Temporal Graphs
Jun 26, 2026Aniq Ur RahmanTemporal Graph Neural Networks Logistic Regression
Department of Engineering Science, University of Oxford, Oxford, OX1 3PJ, UK

Sep 15, 2026, cs.LGJ/K move · Enter open · S save

EdgeReMIND: A Scalable, Top-Ranked Memorization Baseline for Temporal Multi-Relational Link Prediction

Bryant Pollard

Clemson University

Zero-Shot Link Prediction Memo

Open paper AskPDF arXiv

Temporal link prediction on the Temporal Graph Benchmark 2.0 (TGB 2.0) faces a scalability ceiling: on the benchmark's three largest datasets, every existing embedding method runs out of memory or exceeds the time budget. These large-scale graphs are the ones nearest real deployment scale, so failing on them is a real production limitation. EdgeReMIND sets the highest reported test mean reciprocal rank (MRR) on six of eight TGB 2.0 datasets and is the only relation-aware method that runs on all of them. This linear memorization model, with learned per-relation weights over data-calibrated features, is therefore not merely a fallback where embeddings fail but a practical state-of-the-art baseline across the benchmark.

Abstract

Explore similar work

EdgeReMIND: A Scalable, Top-Ranked Memorization Baseline for Temporal Multi-Relational Link Prediction

Where Does Neural Advantage Arise in Continuous-Time Dynamic Graph Prediction?

Estimation-Prediction Tradeoff in Causal Probabilistic Temporal Graphs