{"id":"3dbd8675-66eb-49fa-9743-030148a9597a","arxiv_id":"2606.26145","paper_version":1,"verdict":"UNVERDICTED","confidence":"UNKNOWN","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"NetPTR achieves edge-DP spectral clustering for ordinary networks and column-node-DP for bipartite networks, with consistency guarantees separating non-private error from privacy error under degree-corrected block models in sparse regimes.","lead":"The paper introduces NetPTR, a procedure that adds calibrated noise to the spectral embedding of a network after a stability test to achieve differential privacy while recovering communities. This approach could support analysis of sensitive relational data such as social or biological networks without exposing individual connections.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"No significant objection identified","rationale":"The reader's weakest_assumption correctly isolates the perturbation bounds as the hinge. With the full text now referenced, that step is treated as supplied rather than missing, so the UNVERDICTED status is not altered by an additional concern. No other load-bearing gap (e.g., in the lower-bound argument or bipartite extension) rises to the same level of centrality for the ordinary-network claim.","tokens_in":1777,"tokens_out":324,"duration_ms":18708,"concrete_test":"Extract the explicit perturbation bound (likely in the section deriving stability certificates) and substitute the sparse DC-SBM parameters (average degree ~ log n, community separation constant) to confirm the local sensitivity remains O(1) or polylog(n) rather than diverging; recompute the resulting privacy noise scale and check whether the total error still separates as claimed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on deriving or invoking perturbation bounds for empirical eigenspaces that are tight enough to separate non-private spectral error from privacy noise while remaining computable for local sensitivity in the sparse DC-SBM regime. The abstract states that the algorithm requires such bounds and that they yield stability certificates; the full text presumably supplies the derivation. No internal inconsistency or unsupported step is visible from the provided description, and the matching lower bound on privacy budget is presented as evidence of sharpness. Because the full manuscript is referenced as available, the perturbation-bounds step appears to be the intended technical core rather than an unexamined assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper proposes NetPTR, a private spectral clustering procedure for community detection in ordinary and bipartite networks under differential privacy. For ordinary networks, it achieves edge differential privacy and provides an error bound under the degree-corrected stochastic blockmodel that separates the non-private spectral clustering error from the privacy-induced error, guaranteeing weak consistency in sparse networks and exact recovery in moderately sparse networks, with a matching lower bound on the privacy budget. A similar algorithm is developed for bipartite networks with column-node-DP and consistency under a bipartite degree-corrected block model. Simulations and real-data examples are used to illustrate the privacy-accuracy tradeoff.","tokens_in":1874,"tokens_out":362,"duration_ms":26178,"significance":"If the claimed perturbation bounds for empirical eigenspaces under neighboring-network changes hold and yield computable stability certificates and local sensitivity bounds, the work supplies a theoretically grounded method for differentially private community detection that explicitly decomposes statistical and privacy errors in the sparse DC-SBM regime. The matching lower bound on the privacy budget and the extension to bipartite networks with column-node-DP strengthen the optimality and generality claims.","major_comments":[],"minor_comments":[{"comment":"The abstract states that the algorithm 'requires perturbation bounds... which yield computable stability certificates'; the main text should include an explicit statement (e.g., in the algorithm description or §3) of where these bounds are proved and how they are used to set the noise scale.","section":null},{"comment":"Notation for the privacy budget (ε, δ) and the local sensitivity quantities should be introduced consistently in the introduction and reused without redefinition in the consistency theorems.","section":null}],"recommendation":"minor_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the positive summary, significance assessment, and recommendation of minor revision. No specific major comments were listed in the report, so we have no points to address point-by-point. We will incorporate any minor editorial suggestions during revision.","responses":[],"tokens_in":1277,"tokens_out":68,"duration_ms":11073,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"Hi,\n\nThe one thing to know about this paper is that it introduces NetPTR, which adds a stability test before releasing a noisy spectral embedding under differential privacy, and proves it achieves weak consistency in sparse degree-corrected blockmodels with a matching lower bound on the privacy budget.\n\nWhat stands out as new is the use of column-node differential privacy for bipartite networks and the way they separate the usual spectral error from the added privacy noise. The matching lower bound is useful because it indicates the privacy requirement can't be improved much. The paper handles both ordinary edge-DP and the bipartite case, which broadens the applicability.\n\nThe work is done well in the sense that it provides explicit consistency rates for weak and exact recovery in different sparsity levels. The algorithm description ties the stability certificates directly to the perturbation analysis.\n\nThe main soft spot is whether the perturbation bounds for the eigenspaces under single edge or column changes are sufficiently tight in the sparse regime to make the stability test practical. The abstract says they yield computable bounds, but if they require the degrees to be higher or introduce extra logarithmic terms, the guarantees could weaken. Without the full proofs it's difficult to assess if everything lines up without circularity or extra assumptions.\n\nThis is aimed at people working on privacy for relational data in networks. Someone interested in theoretical DP methods for community detection would find the rates and lower bound worth looking at. It has enough formal content to merit a serious referee, even if revisions are needed on the bound derivations.\n\nI would recommend sending it for peer review.","headline":"NetPTR adds stability testing to private spectral community detection with consistency guarantees and a sharp lower bound.","tokens_in":2374,"tokens_out":381,"would_cite":false,"duration_ms":31277,"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":"NetPTR releases a noisy spectral embedding after a stability test to achieve edge differential privacy while preserving consistency guarantees for community detection.","keywords":["differential privacy","spectral clustering","community detection","stochastic blockmodel","sparse networks","edge differential privacy","bipartite networks"],"falsifier":"A sparse network drawn from the degree-corrected stochastic blockmodel in which the observed clustering error after privacy noise exceeds the derived upper bound, or in which the stability test passes on an embedding whose neighboring-network change is large.","tokens_in":2655,"feed_emoji":"🔒","tokens_out":683,"duration_ms":18370,"temperature":0.7,"pith_summary":"The paper introduces NetPTR as a procedure for spectral community detection that adds calibrated noise to the empirical embedding only after a stability test passes. For ordinary networks this yields edge differential privacy together with explicit error bounds under the degree-corrected stochastic blockmodel. The bounds isolate the usual non-private clustering error from the extra term caused by privacy noise, which is enough to retain weak consistency on sparse networks and exact recovery on moderately sparse ones. A matching lower bound shows that the privacy budget needed cannot be improved by more than logarithmic factors. The same framework is extended to column-node differential privacy on bipartite networks with analogous consistency results under a bipartite degree-corrected block model.","feed_headline":"Edge-DP spectral clustering preserves consistency in sparse networks","feed_subtitle":"Error bounds separate privacy cost from clustering error, with matching lower bound on required budget.","key_machinery":"The stability test on empirical eigenspaces that yields computable stability certificates and local sensitivity bounds, allowing calibrated noise to be added to the spectral embedding while controlling the privacy-accuracy tradeoff.","core_discovery":"NetPTR achieves edge differential privacy for ordinary networks by releasing a noisy empirical spectral embedding after a stability test, establishing error bounds under the degree-corrected stochastic blockmodel that separate non-private spectral clustering error from additional privacy error, guaranteeing weak consistency in sparse networks and exact recovery in moderate sparse networks, with a matching lower bound showing the privacy budget is sharp up to log factors. It also extends to column-node-DP for bipartite networks under bipartite degree-corrected block model.","pith_inferences":["The same stability-test-plus-noise pattern could be applied to other eigenvector-based graph algorithms such as spectral ranking.","Networks whose degree distribution deviates strongly from the model assumptions would require separate verification of the perturbation bounds.","The clean separation of error terms suggests that privacy overhead becomes negligible once average degree exceeds a moderate threshold.","Adaptive choice of the stability threshold on real data could be tested by monitoring how often the test rejects valid embeddings."],"forward_implications":["Weak consistency holds for community detection on sparse networks under edge differential privacy.","Exact recovery becomes possible on moderately sparse networks once the privacy noise is added.","The privacy budget required is information-theoretically tight up to logarithmic factors.","A column-node differential privacy version yields consistent recovery for the left-side communities in bipartite networks."],"fun_headline_variants":["NetPTR adds edge-DP to sparse network spectral clustering","Privacy error separated from clustering error in edge-DP","Matching lower bound sharpens privacy budget for NetPTR","Column-node-DP extends NetPTR to bipartite networks"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"Perturbation bounds for empirical eigenspaces under neighboring-network changes must hold so that stability certificates and local sensitivity bounds can be computed.","fun_headline_variants_meta":{"raw":{"variants":["NetPTR adds edge-DP to sparse network spectral clustering","Privacy error separated from clustering error in edge-DP","Matching lower bound sharpens privacy budget for NetPTR","Column-node-DP extends NetPTR to bipartite networks"]},"model":"grok-4.3","cost_usd":0.009262,"raw_usage":{"total_tokens":4157,"prompt_tokens":689,"num_sources_used":0,"completion_tokens":63,"cost_in_usd_ticks":92624500,"prompt_tokens_details":{"text_tokens":689,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":3405,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":689,"tokens_out":63,"duration_ms":20939,"temperature":1.0,"reasoning_tokens":3405,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-26T06:36:02.131267+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A sparse network drawn from the degree-corrected stochastic blockmodel in which the observed clustering error after privacy noise exceeds the derived upper bound, or in which the stability test passes on an embedding whose neighboring-network change is large.","supporting_citations":[],"review_version":1}