{"id":"809c5bf0-f3c3-46eb-8a19-0a6328aa63f7","arxiv_id":"2605.28155","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"HERMIT combines hyperbolic TGNN with RTT-aware edges and random forest regression to deliver 6% RMSE gain over historical baselines on 2015-2024 internet data while improving link prediction over prior hyperbolic models.","lead":"The paper proposes HERMIT, a hybrid model using a hyperbolic temporal graph neural network plus random forest to predict internet round-trip times on real routing data. A smart generalist might read it for insight into using non-Euclidean geometry to handle hierarchical network structures in practical latency forecasting.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.3","headline":"Missing ablations leave unclear if hyperbolic TGNN drives the 6% RMSE gain over historical-RF baseline","rationale":"The reader's weakest assumption directly matches the missing control needed to attribute gains to the hyperbolic temporal GNN rather than to extra learned features. This is the single most load-bearing point for the outperformance claim; other aspects (dataset scale, hybrid design) are secondary until this is resolved.","tokens_in":1767,"tokens_out":337,"duration_ms":43607,"concrete_test":"Re-train the hybrid pipeline with the hyperbolic GNN replaced by an otherwise identical Euclidean TGNN (same edge encoder, temporal layers, and RF head) on the identical 2015-2024 split; if the RMSE gap versus the historical-RF baseline falls below 3% or loses significance on heavy-tailed quantiles, the hyperbolic component is not load-bearing for the headline claim.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim requires that the manifold-preserving temporal GNN with RTT-aware edge encoder produces node representations that, when fed to the RF alongside historical statistics, yield a genuine 6% RMSE reduction (and fewer heavy-tail errors) beyond what the RF achieves on historical RTT alone. Because the final predictor is still an RF on concatenated features, the improvement could arise from any additional signal in the GNN embeddings rather than from hyperbolic geometry or temporal manifold preservation. The abstract reports superiority only versus prior hyperbolic models (HMPTGN, HTGN) for link prediction and versus a pure historical-RF baseline for RTT; no Euclidean TGNN or GNN-ablated control is mentioned.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript proposes HERMIT, a hybrid model that augments the HMPTGN hyperbolic temporal GNN with an RTT-aware learnable edge encoder; the resulting node embeddings are concatenated with historical RTT statistics and fed to a Random Forest regressor for joint link prediction and RTT regression on Internet-scale graphs. On a 2015-2024 real-world dataset the method is reported to yield a 6% RMSE reduction versus a pure historical-RTT Random Forest baseline while also outperforming prior hyperbolic TGNNs (HMPTGN, HTGN) on link prediction.","tokens_in":1904,"tokens_out":593,"duration_ms":25899,"significance":"If the reported gains are shown to arise specifically from the hyperbolic manifold-preserving component rather than from generic additional features, the work would supply a concrete, scalable demonstration that hyperbolic TGNNs can improve latency modeling on heavy-tailed, scale-free routing graphs beyond what Euclidean TGNNs or simple statistical baselines achieve.","major_comments":[{"comment":"The central empirical claim (6% RMSE improvement over the historical-RF baseline) is load-bearing for the paper's contribution, yet the evaluation section provides no ablation that replaces the hyperbolic TGNN with an otherwise identical Euclidean TGNN or with a non-manifold-preserving temporal GNN; without this control it remains possible that any additional embedding signal, rather than hyperbolic geometry or manifold preservation, drives the reported gain.","section":"Experiments / Evaluation"},{"comment":"Table reporting RTT prediction results (presumably Table X) shows only aggregate RMSE; the claim that large errors on heavy-tailed samples are reduced is not supported by a quantile-specific or tail-specific metric, nor by a statistical test against the baseline, making it impossible to verify the heavy-tail improvement asserted in the abstract.","section":"Experiments / RTT Prediction Results"}],"minor_comments":[{"comment":"Dataset description is incomplete: no statistics on number of nodes, edges, temporal snapshots, or RTT distribution moments are supplied, preventing assessment of scale or representativeness.","section":"Dataset"},{"comment":"No error bars, standard deviations across runs, or cross-validation details are reported for either the link-prediction or RTT-prediction tasks.","section":"Experiments"},{"comment":"Notation for the learnable edge encoder and the manifold-preserving temporal aggregation is introduced without an explicit equation or pseudocode block, making the architectural novelty hard to reproduce from the text alone.","section":"Method"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is submitted to a machine-learning venue yet centers on an applied networking task; the editor may wish to confirm topical fit. The absence of any Euclidean control is the primary technical gap; once addressed the paper could be publishable with only minor polishing."},"author_rebuttal":{"model":"grok-4.3","summary":"Thank you for the referee's constructive comments. We agree that the suggested additions will strengthen the empirical support for our claims and will incorporate them in the revised manuscript.","responses":[{"response":"We agree that an ablation isolating the hyperbolic manifold's contribution is required. In the revised manuscript we will add this control by implementing an otherwise identical Euclidean TGNN (replacing all hyperbolic operations with their Euclidean counterparts while retaining the same architecture, RTT-aware edge encoder, and Random Forest integration) and reporting its performance on both link prediction and RTT regression relative to HERMIT and the historical baseline.","revision_made":"yes","referee_comment":"[Experiments / Evaluation] The central empirical claim (6% RMSE improvement over the historical-RF baseline) is load-bearing for the paper's contribution, yet the evaluation section provides no ablation that replaces the hyperbolic TGNN with an otherwise identical Euclidean TGNN or with a non-manifold-preserving temporal GNN; without this control it remains possible that any additional embedding signal, rather than hyperbolic geometry or manifold preservation, drives the reported gain."},{"response":"We acknowledge that aggregate RMSE alone does not substantiate the heavy-tail claim. In revision we will augment the RTT prediction results with quantile-specific metrics (RMSE and MAE at the 75th, 90th, and 95th percentiles of the error distribution) for HERMIT versus the baseline, together with a statistical test (Wilcoxon signed-rank) on the per-sample errors restricted to the upper tail of the RTT distribution.","revision_made":"yes","referee_comment":"[Experiments / RTT Prediction Results] Table reporting RTT prediction results (presumably Table X) shows only aggregate RMSE; the claim that large errors on heavy-tailed samples are reduced is not supported by a quantile-specific or tail-specific metric, nor by a statistical test against the baseline, making it impossible to verify the heavy-tail improvement asserted in the abstract."}],"tokens_in":1437,"tokens_out":418,"duration_ms":34517,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"HERMIT gets a 6% RMSE improvement on RTT prediction over a historical statistics baseline by feeding in embeddings from a hyperbolic temporal GNN with RTT-aware edges. The paper also claims better link prediction than previous hyperbolic models like HMPTGN and HTGN.\n\nThe work applies this to a decade-long real internet topology dataset, which is a plus for relevance to actual routing problems. Adding learnable edge encoders for RTT features is a reasonable extension to handle evolving link states in scale-free graphs.\n\nThe main weakness is the lack of controls that would show the hyperbolic manifold or the temporal preservation is what drives the gains. The final model is still an RF on top of the GNN outputs plus history, so any extra signal from the GNN could explain the difference. No comparison to a Euclidean TGNN appears, and no ablation removing the manifold-preserving aspects. That leaves open whether a simpler graph model would suffice. The heavy-tailed error reduction is noted but without supporting numbers or tests.\n\nThe paper is for people working on graph-based methods for network performance. Someone looking at practical applications in traffic engineering could get value from the dataset handling and the hybrid setup. The thinking seems straightforward and engaged with the relevant prior TGNN work.\n\nI would send this to peer review. The real-world data and the concrete task make it worth referee time, though the authors should add the missing ablations to strengthen the claims.","headline":"HERMIT gets a modest 6% RMSE edge on RTT prediction by adding hyperbolic TGNN embeddings to an RF, but the experiments leave it unclear whether the hyperbolic geometry is what matters.","tokens_in":2385,"tokens_out":369,"would_cite":false,"duration_ms":32677,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"Hyperbolic temporal GNN with RTT-aware edges and random forest regression improves RTT prediction by 6% RMSE over historical statistics alone on real internet data.","keywords":["hyperbolic graph neural networks","temporal graph neural networks","RTT prediction","internet routing","delay prediction","scale-free networks","link prediction","heavy-tailed distributions"],"falsifier":"A direct replication on the same 2015-2024 dataset in which HERMIT shows no RMSE improvement, no reduction in large errors on heavy-tailed samples, or no link-prediction gains relative to the random forest baseline using only historical RTT statistics.","tokens_in":2673,"feed_emoji":"🌐","tokens_out":709,"duration_ms":27070,"temperature":0.7,"pith_summary":"The paper seeks to establish that internet routing graphs, being scale-free and hierarchical, are better modeled in hyperbolic space than Euclidean space for capturing temporal evolution and predicting round-trip times. It introduces HERMIT as a hybrid that augments a hyperbolic manifold-preserving temporal GNN with a learnable RTT-aware edge encoder and feeds the node representations into a random forest regressor alongside historical statistics. Evaluation on a 2015-2024 real-world dataset shows consistent gains in both RTT accuracy, especially on heavy-tailed samples, and link prediction over prior hyperbolic TGNNs and pure statistical baselines. A sympathetic reader would care because RTT prediction directly affects routing decisions, quality-of-service guarantees, and traffic engineering in operational networks. If the claim holds, the work demonstrates a practical way to combine geometric graph learning with tree-based regression for delay forecasting.","feed_headline":"Hyperbolic GNN plus random forest cuts RTT error by 6%","feed_subtitle":"Model captures scale-free routing structure to reduce large errors on real 2015-2024 internet data beyond historical statistics alone.","key_machinery":"Hyperbolic manifold-preserving temporal GNN augmented by a learnable RTT-aware edge encoder that produces node representations for joint link and latency prediction.","core_discovery":"HERMIT integrates a hyperbolic manifold-preserving temporal GNN built on HMPTGN with RTT-aware edge features and a learnable edge encoder, then combines the resulting node representations with historical RTT statistics inside a random forest regressor; on a large real internet dataset this hybrid yields a 6% RMSE reduction versus a strong random forest baseline that uses only historical statistics, fewer large errors on heavy-tailed samples, and better link prediction than earlier hyperbolic TGNNs such as HMPTGN and HTGN.","pith_inferences":["The same hyperbolic-plus-regressor pattern could be tested on other scale-free networks such as social or citation graphs to check whether the geometry advantage generalizes.","Directly feeding the predicted RTT values into existing routing protocols might produce measurable gains in end-to-end path selection that the paper does not evaluate.","Training the model on shorter time slices or different geographic subsets of the data would test whether the reported gains remain stable under distribution shift.","An end-to-end differentiable replacement for the random forest component could be compared to isolate how much of the gain comes from the hyperbolic representations versus the hybrid architecture."],"forward_implications":["Better modeling of long-term temporal dependencies and evolving routing dynamics in internet topologies.","Reduced prediction errors on heavy-tailed latency distributions for more reliable QoS provisioning.","Improved link prediction accuracy within the same hyperbolic framework.","A scalable hybrid pipeline that combines geometric graph learning with tree-based regression for real-world delay forecasting."],"fun_headline_variants":["Hyperbolic GNN plus RF cuts RTT error 6%","HERMIT improves RTT prediction 6% via hyperbolic GNN","Scale-free internet modeled better with temporal hyperbolic graphs","HERMIT reduces RMSE 6% over historical statistics baseline"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"Hyperbolic geometry plus the learnable RTT-aware edge encoder will capture evolving link states and routing behavior in scale-free internet graphs more effectively than Euclidean TGNNs or historical statistics alone.","fun_headline_variants_meta":{"raw":{"variants":["Hyperbolic GNN plus RF cuts RTT error 6%","HERMIT improves RTT prediction 6% via hyperbolic GNN","Scale-free internet modeled better with temporal hyperbolic graphs","HERMIT reduces RMSE 6% over historical statistics baseline"]},"model":"grok-4.3","cost_usd":0.007207,"raw_usage":{"total_tokens":3358,"prompt_tokens":735,"num_sources_used":0,"completion_tokens":64,"cost_in_usd_ticks":72074500,"prompt_tokens_details":{"text_tokens":735,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":2559,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":735,"tokens_out":64,"duration_ms":23496,"temperature":1.0,"reasoning_tokens":2559,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-29T14:06:18.454293+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A direct replication on the same 2015-2024 dataset in which HERMIT shows no RMSE improvement, no reduction in large errors on heavy-tailed samples, or no link-prediction gains relative to the random forest baseline using only historical RTT statistics.","supporting_citations":[],"review_version":1}