{"id":"08714c5b-ec44-4f98-a15f-4b463d4c0bd7","arxiv_id":"2604.15748","paper_version":3,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"CoAt-CBM uses concept-wise visual queries and concept contrastive optimization to improve fine-grained image–concept alignment in CLIP-based concept bottleneck models.","lead":"The paper proposes CoAt-CBM, a concept bottleneck model that uses learnable concept-wise visual queries and a concept contrastive loss to get finer image–concept alignment than prior CLIP-based CBMs. If the gains hold, it would be a practical upgrade for interpretable image classifiers that need concept-level explanations.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.5","headline":"Manuscript body is an unrelated contact-process paper; CoAt-CBM claims have zero supporting text, equations, or experiments.","rationale":"The reader correctly diagnosed the abstract/body mismatch and set UNVERDICTED with low confidence. The single load-bearing failure is not a subtle assumption inside CoAt-CBM but the total absence of any CoAt-CBM content: the provided “full manuscript” is arXiv 2604.15753 (long-range contact process). Without methods, loss derivation, or results, no technical claim—contrastive relative-importance, adaptive queries, or SOTA outperformance—can be checked. This is stronger than the reader’s weakest-assumption note; the assumption cannot be evaluated at all. Verdict remains UNVERDICTED; if the correct PDF is later supplied the review must restart from scratch. No other internal inconsistency can be assessed because the relevant text is missing.","tokens_in":32621,"tokens_out":458,"duration_ms":10567,"concrete_test":"Retrieve the actual PDF of arXiv:2604.15748 and confirm whether its body contains any CoAt-CBM equations, architecture diagram, or experimental tables; if it instead matches the contact-process text (or is empty), the central claim is unsupported and the paper cannot be scored.","verdict_should_be":"UNVERDICTED","load_bearing_attack":"The abstract asserts that CoAt-CBM’s learnable concept-wise visual queries plus “novel concept contrastive optimization” overcome CLIP granularity/structural biases and BCE’s independent-concept treatment, yielding faithful alignment and SOTA gains. The supplied full manuscript, however, is entirely different: “The truncation property and continuity for the long-range contact process on Zd” (Theorems 1.2–1.4, renormalization arguments, graphical constructions, Propositions 4.3–5.2). No architecture, loss formula, concept-score vector, contrastive objective, ablation, or table for any CBM appears. Consequently the strongest claim rests on an empty document; the mechanism the reader flagged cannot even be inspected, let alone validated.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The submission metadata and abstract describe CoAt-CBM, a Concept Bottleneck Model that uses learnable concept-wise visual queries to extract fine-grained concept embeddings and a concept contrastive objective (instead of independent BCE) to enforce relative concept importance, claiming adaptive image–concept alignment, high interpretability, and consistent SOTA gains over CLIP-based CBMs. The body of the manuscript, however, is an entirely different paper: “The truncation property and continuity for the long-range contact process on Zd” (Bethuelsen & Namugera), which develops renormalization arguments, graphical constructions, and Theorems 1.2–1.4 on resilience/truncation and continuity of the survival probability for long-range contact processes. No CoAt-CBM architecture, loss, concept-score definition, experiment, dataset, or ablation appears anywhere in the full text.","tokens_in":32780,"tokens_out":967,"duration_ms":16133,"significance":"If the abstract’s claims were supported by a matching manuscript, a method that mitigates CLIP granularity bias and mutual-exclusivity failures in CBMs would be of clear interest to the interpretability and vision communities. As submitted, the CoAt-CBM contribution cannot be assessed: there is no method description, no formal statement of the contrastive objective, and no empirical evidence. The attached contact-process results are a solid contribution to interacting particle systems, but they are outside the scope of a cs.CV venue and do not substantiate the abstract. Significance of the claimed CBM work is therefore zero on the present document.","major_comments":[{"comment":"Title/abstract vs. full text mismatch: the abstract and paper_id claim CoAt-CBM (cs.CV), but the entire manuscript body (arXiv-style header, AMS-MSC 60K35/82B43, Theorems 1.2–1.4, Sections 3–6, graphical construction of the LRCP, Propositions 4.3–5.2, renormalization) is a probability paper on long-range contact processes. No concept bottleneck, CLIP, attention query, or contrastive loss is defined. The central CoAt-CBM claims are therefore unsupported by any technical content.","section":null},{"comment":"Absence of the claimed method: the abstract asserts that “learnable concept-wise visual queries” produce a concept score vector and that “concept contrastive optimization” handles relative importance and mutual exclusivity. None of these objects—queries, score vector, contrastive loss formula, training procedure, or interpretability metric—appear in any section or equation of the supplied manuscript.","section":null},{"comment":"Absence of experiments: the abstract states “Extensive experiments demonstrate that CoAt-CBM consistently outperforms state-of-the-art methods.” The manuscript contains no datasets, metrics, tables, ablations, baselines, or statistical tests related to CBMs (or to any vision task). The performance claim cannot be evaluated.","section":null},{"comment":"Scope and venue fit: even if the contact-process theorems (extinction at criticality for resilient λ, continuity of φ under (1.5), truncation for α > d symmetric / α > 2d+1 non-symmetric) are correct, they belong in a probability journal, not a cs.CV venue reviewing Concept Bottleneck Models. The submission as packaged is not reviewable for the stated contribution.","section":null}],"minor_comments":[{"comment":"If this is a packaging/upload error (wrong PDF attached to the CoAt-CBM abstract), the authors should resubmit the correct manuscript; the present document cannot be revised into a CBM paper within the same file.","section":null},{"comment":"The contact-process manuscript itself has standard presentation (clear theorems, graphical construction in §3, renormalization in §§4–5); those issues are irrelevant to the CoAt-CBM review.","section":null}],"recommendation":"reject","confidential_remarks":"This appears to be a severe metadata/body mismatch (cs.CV abstract for CoAt-CBM paired with a math.PR contact-process paper, arXiv 2604.15753-style content). Likely an automated or human upload error rather than intentional fraud, but the submission is not reviewable as a CBM paper. Recommend desk-reject / return to authors for correct PDF; do not send the contact-process text to CV reviewers. If the journal also handles probability, the contact-process paper could be redirected separately."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The one thing you need to know: the abstract and the manuscript do not match. The abstract sells CoAt-CBM—learnable concept-wise visual queries plus a concept contrastive loss on top of CLIP-style CBMs, with a SOTA claim. The full text is Bethuelsen & Namugera on truncation and continuity for long-range contact processes on Zd (renormalization, graphical construction, Theorems 1.2–1.4). Zero architecture, loss formula, concept scores, ablations, or vision experiments appear.\n\nSo for CoAt-CBM there is nothing new we can credit beyond a short abstract. The two ingredients named there (concept-wise queries; relative/contrastive concept scoring instead of independent BCE) are a plausible incremental CBM recipe, but they are only asserted. We cannot check whether they fix CLIP granularity bias or mutual exclusivity, or whether the SOTA claim holds.\n\nThe contact-process manuscript, taken on its own terms, looks like careful probability work: clear resilience/truncation definitions, finite space-time conditions adapted from Bezuidenhout–Grimmett/Gray-style renormalization to long-range rates, and continuity of the survival probability under resilient parameters. That is real math for a different arXiv id. It does not support the CBM paper.\n\nSoft spot is not a weak experiment section—it is an empty one for the claimed contribution. Citation pattern and empirical design for CoAt-CBM are uninspectable. Circularity or overclaim cannot even be stress-tested.\n\nWho is this for? Nobody, until the correct CoAt-CBM PDF is attached. A serious editor should desk-reject or bounce for the wrong manuscript, not burn referee time on a title/body mismatch. I would not bring this package to reading group, would not cite CoAt-CBM from this material, and would not send it to peer review as submitted. If the real CoAt-CBM paper arrives with methods, ablations, and tables, re-open; until then there is no paper to engage.","headline":"Abstract is a CLIP-CBM methods claim; the supplied body is an unrelated long-range contact-process paper, so CoAt-CBM cannot be evaluated at all.","tokens_in":33412,"tokens_out":530,"would_cite":false,"duration_ms":12986,"reading_group":"no","serious_thinker":"unclear","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60K35","82B43"],"pacs":[],"model":"grok-4.5","headline":"Long-range contact processes that stay supercritical after truncation of far-away infections also die out exactly at criticality, and their survival probability varies continuously with the rates.","keywords":["contact process","long-range interactions","renormalization","truncation property","continuity of critical value","resilience","interacting particle systems"],"falsifier":"Construct a concrete summable, translation-invariant infection rate on Z^2 whose power-law exponent lies between d and 2d+1, for which either truncation at every finite range drives the survival probability to zero while the untruncated process survives, or the survival probability jumps discontinuously under a continuous change of rates.","tokens_in":33463,"feed_emoji":"🔗","tokens_out":656,"duration_ms":17994,"temperature":0.7,"pith_summary":"The paper studies contact processes on the integer lattice that allow infection to jump arbitrarily far, provided the infection rates are summable and translation-invariant. It shows that when the rates decay fast enough (roughly power-law with exponent larger than the dimension, or a bit stronger without symmetry), a supercritical process remains supercritical even after all infections longer than some large but finite distance are deleted. For every such “resilient” family of rates the process dies out at the critical recovery rate, and the probability of never recovering is continuous under natural perturbations of the rates. The results extend the classical finite-range theory of Bezuidenhout–Grimmett by adapting renormalization arguments to unbounded interactions, giving a practical criterion under which long-range models behave like their truncated finite-range cousins.","feed_headline":"Long-range infections stay alive after you cut far jumps","feed_subtitle":"Power-law contact processes remain supercritical under truncation and have continuous survival probability","key_machinery":"Resilience (truncation property) of the infection parameter together with a finite space-time condition obtained by renormalization: with high probability an infected finite set A reappears, after a controlled time, both inside a large box and on its lateral boundaries. This condition is proved separately for symmetric power-law rates (alpha > d) and for non-symmetric rates with faster decay (alpha > 2d+1), then fed into the classical renormalization comparison with supercritical oriented percolation.","core_discovery":"If the infection rates of a long-range contact process on Zd are resilient (they satisfy the truncation property), then the process is critical if and only if the expected space-time cluster is infinite while the survival probability is zero; moreover the survival probability is continuous in both the infection rates and the recovery rate whenever the rates remain resilient.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Resilient infections stay critical after truncating long jumps","Power-law contact processes remain supercritical under truncation","Truncation preserves criticality when rates have the resilience property","Survival probability stays continuous for resilient infection rates","Infinite clusters mark criticality in truncated long-range processes"],"cache_read_input_tokens":16512,"weakest_assumption_plain":"The infection rates must decay at least as fast as a power law with exponent larger than the dimension (or 2d+1 without symmetry); slower decay may destroy both resilience and the linear-growth bound used in the proofs.","fun_headline_variants_meta":{"raw":{"variants":["Resilient infections stay critical after truncating long jumps","Power-law contact processes remain supercritical under truncation","Truncation preserves criticality when rates have the resilience property","Survival probability stays continuous for resilient infection rates","Infinite clusters mark criticality in truncated long-range processes"]},"model":"grok-4.5","effort":"low","cost_usd":0.00573,"raw_usage":{"total_tokens":1523,"prompt_tokens":756,"num_sources_used":0,"completion_tokens":77,"cost_in_usd_ticks":57300000,"prompt_tokens_details":{"text_tokens":756,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":690,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":756,"tokens_out":77,"duration_ms":5260,"temperature":1.0,"reasoning_tokens":690,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-12T19:34:56.488847+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Construct a concrete summable, translation-invariant infection rate on Z^2 whose power-law exponent lies between d and 2d+1, for which either truncation at every finite range drives the survival probability to zero while the untruncated process survives, or the survival probability jumps discontinuously under a continuous change of rates.","supporting_citations":[],"review_version":2}