{"id":"6cdb6b0f-341b-469e-a374-8e4891a2f3df","arxiv_id":"2605.14727","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":2,"one_line_summary":"CHASM introduces a cross-frequency harmonized axis-separable spectral mixer using a shared channel eigenbasis plus per-frequency positive gains, yielding consistent gains over same-backbone baselines in medical and natural image tasks.","lead":"CHASM proposes a spectral token mixer that shares one learned channel eigenbasis across all frequencies while allowing each frequency its own positive gain factors, then applies the operator separably along image height and width. This structured design is positioned as a drop-in replacement that improves performance on MRI reconstruction, MRI segmentation, and natural-image tasks over prior spectral mixers.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"Uncontrolled parameter counts could explain gains; shared eigenbasis benefit may be capacity artifact rather than inductive bias","rationale":"The reader's weakest assumption is identical to the identified concern. The abstract already gestures at ablations, but without explicit parameter matching those ablations cannot rule out capacity confounds. This is the single most load-bearing point for the 'consistently improves' claim; all other elements (structural characterization, axis-separable application) are secondary once capacity is controlled.","tokens_in":1723,"tokens_out":350,"duration_ms":31222,"concrete_test":"From the experimental sections, extract exact parameter counts for CHASM and each same-backbone spectral-mixer baseline on at least one task (e.g., accelerated MRI reconstruction). If counts differ by >5%, re-train all models with width/depth adjustments to enforce exact parameter parity and re-measure the performance delta; if the CHASM advantage disappears or reverses, the inductive-bias claim does not hold.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is that the shared channel eigenbasis supplies a useful inductive bias, evidenced by consistent improvements and ablations where removing the shared-basis constraint weakens performance. However, the operator description (shared eigenbasis + per-frequency positive gains, applied axis-separably) does not specify whether this construction matches the parameter count of the frequency-indexed baselines. If CHASM uses more total parameters (e.g., via the shared basis plus gains), the reported gains could be explained by capacity rather than the cross-frequency harmonization. The abstract's ablation claim is load-bearing only if the ablated variants are parameter-matched; otherwise the evidence does not isolate the inductive bias.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper proposes CHASM, a spectral token mixer for visual feature maps that enforces a shared learned channel eigenbasis across frequencies while allowing per-frequency positive spectral gains, applied axis-separably along height and width. It is positioned as a drop-in replacement in existing backbones and is evaluated via controlled same-backbone comparisons on accelerated MRI reconstruction, undersampled MRI segmentation, and natural-image reconstruction tasks, where it reports consistent gains over spectral-mixer baselines. Ablations are cited to show that removing the shared-basis constraint weakens performance and that randomizing sampling geometry reduces the benefit, framing the shared eigenbasis as a useful inductive bias for cross-frequency harmonization.","tokens_in":1870,"tokens_out":559,"duration_ms":18627,"significance":"If the reported gains prove robust after parameter-matched controls and statistical verification, CHASM would supply a concrete structural prior for spectral operators that separates shared channel directions from frequency-specific scaling. This could be useful in domains like medical imaging where global frequency interactions matter and where existing Fourier-based mixers lack explicit cross-frequency alignment. The structural characterization of the shared-basis family is a positive element that could support future analysis.","major_comments":[{"comment":"Abstract and operator description: the central claim that the shared channel eigenbasis supplies a useful inductive bias (rather than a capacity artifact) rests on same-backbone comparisons and ablations, yet no statement confirms that CHASM and the frequency-indexed baselines have identical parameter counts. The construction (shared eigenbasis plus per-frequency gains) appears to add parameters relative to purely frequency-indexed mixing; without explicit matching or an ablation that isolates the basis while holding total parameters fixed, the performance delta cannot be attributed to cross-frequency harmonization.","section":"Abstract"}],"minor_comments":[{"comment":"The abstract states 'consistent improvements' and 'ablations show' but supplies no quantitative values, error bars, exact baseline implementations, data splits, or statistical tests; these details are required to assess robustness.","section":"Abstract"},{"comment":"The positive spectral gains are described as 'positive' but the precise constraint (e.g., ReLU, softplus, or projection) and its effect on the operator's spectral properties should be stated explicitly.","section":"Method"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's empirical claims are presented at a high level with no numerical results or implementation details visible in the provided abstract, which lowers confidence in the soundness assessment. The parameter-count concern is the primary load-bearing issue; if addressed with matched ablations and full experimental protocols, the work could be suitable for a vision or medical-imaging venue."},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the thoughtful review and for identifying the need to clarify parameter counts in our comparisons. We address the concern directly below and will update the manuscript to make the parameter analysis explicit.","responses":[{"response":"We appreciate this observation. In fact, CHASM uses substantially fewer parameters than a purely frequency-indexed mixer. A frequency-indexed baseline applies an independent channel-mixing matrix at each frequency, incurring O(F·C²) parameters. CHASM instead learns one shared eigenbasis (O(C²)) and a scalar positive gain per frequency (O(F)), for a total of O(C² + F) parameters. The reported gains are therefore obtained with a strictly smaller model, which reinforces rather than undermines the value of the shared-basis inductive bias. We will revise the manuscript to (i) report exact parameter counts for CHASM and every baseline in the experimental tables, (ii) add a brief statement in the abstract and operator section confirming the parameter relationship, and (iii) include an additional ablation that inflates the baseline capacity to match or exceed CHASM’s parameter budget.","revision_made":"yes","referee_comment":"[Abstract] Abstract and operator description: the central claim that the shared channel eigenbasis supplies a useful inductive bias (rather than a capacity artifact) rests on same-backbone comparisons and ablations, yet no statement confirms that CHASM and the frequency-indexed baselines have identical parameter counts. The construction (shared eigenbasis plus per-frequency gains) appears to add parameters relative to purely frequency-indexed mixing; without explicit matching or an ablation that isolates the basis while holding total parameters fixed, the performance delta cannot be attributed to cross-frequency harmonization."}],"tokens_in":1390,"tokens_out":367,"duration_ms":35002,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"CHASM defines a spectral token mixer that shares one learned channel eigenbasis across frequencies while keeping strictly positive per-frequency gains, then applies the whole thing separably along height and width. This sits between fixed filter designs and fully adaptive frequency-indexed mixing, and the paper supplies a structural characterization of the resulting operator family. That characterization and the explicit separation of shared versus frequency-specific parts are the concrete new elements. The evaluations replace the mixer inside existing backbones and show gains on accelerated MRI reconstruction, undersampled MRI segmentation, and natural-image reconstruction, with ablations that remove the shared-basis constraint and report weaker results. The design avoids obvious circularity by tying the claimed benefit to external task performance rather than internal fitting metrics. The main soft spot is the parameter-count issue the stress-test note flags. The abstract gives no numbers and does not confirm that the ablated variants or the baseline spectral mixers are matched in total parameters, so it remains possible that the improvements come from extra capacity rather than the cross-frequency harmonization bias. No error bars, exact implementation details, or statistical tests appear in the provided summary either, which leaves the robustness of the “consistent improvements” claim hard to judge. This paper is for people working on efficient Fourier-based vision backbones or medical imaging pipelines that already use spectral token mixers. A reader who wants a new inductive bias for making channel directions comparable across frequencies would get something usable from the operator description and the ablation logic. It deserves a serious referee because the core construction is well-specified and the tasks are relevant, even though the empirical controls will need tightening in revision.","headline":"CHASM gives a clean middle-ground spectral mixer via shared channel eigenbasis plus per-frequency gains, but the reported gains rest on unverified parameter counts and thin experimental details.","tokens_in":2411,"tokens_out":399,"would_cite":false,"duration_ms":27309,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":{"model":"grok-4.3","evidence":[],"headline":"CHASM shared-eigenbasis spectral operator is a practical CV design with no overlap to RS forcing chain or J-cost structures","alignment":"orthogonal","rationale":"The paper's core construction (M(k) = U diag(λ_k) U^T with shared orthogonal U and positive per-frequency λ_k, axis-separable FFT application, and manifold dimension C(C-1)/2 + KC) is a learned mixer for Fourier token mixing in MRI/vision backbones. It contains no J-cost, reciprocal symmetry, golden-ratio identities, 8-tick periodicity, parameter-free constant derivation, or distinction-forcing steps. The positive-spectrum Hermitian motivation is generic linear-algebra hygiene, not an instance of the RS J-cost or recognition-ladder theorems. No RS module (e.g., Cost.FunctionalEquation, Foundation.AlexanderDuality, Constants, Spectra) is paralleled or contradicted.","tokens_in":50772,"confidence":"high","tokens_out":203,"duration_ms":13613,"cache_read_input_tokens":32896,"cache_creation_input_tokens":0},"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"CHASM shares one channel eigenbasis across frequencies while keeping per-frequency positive gains to improve spectral token mixers.","keywords":["spectral token mixers","Fourier transforms","cross-frequency harmonization","MRI reconstruction","image segmentation","axis-separable mixing"],"falsifier":"An experiment in which removing the shared-basis constraint leaves performance unchanged or randomizing coherent sampling geometry eliminates the reported gains.","tokens_in":2616,"feed_emoji":"","tokens_out":533,"duration_ms":28243,"temperature":0.7,"pith_summary":"Spectral token mixers based on Fourier transforms model global interactions efficiently but often fail to align channel directions across frequencies. CHASM provides a middle ground by sharing a learned channel eigenbasis among all frequencies and retaining individual positive spectral gains per frequency. This shared basis makes channel directions comparable across the spectrum while the gains preserve frequency-specific adaptivity. The operator is applied separably along height and width axes as a drop-in replacement inside existing backbones. Controlled same-backbone experiments show consistent gains over prior spectral mixers in accelerated MRI reconstruction, undersampled MRI segmentation, and natural-image reconstruction.","feed_headline":"Shared channel basis across frequencies boosts spectral mixers","feed_subtitle":"CHASM keeps frequency-specific gains but enforces one eigenbasis, lifting MRI and natural-image performance.","key_machinery":"Shared channel eigenbasis spectral operator with per-frequency positive gains, applied separably along spatial axes.","core_discovery":"CHASM separates a shared channel eigenbasis, used by every frequency, from frequency-specific positive spectral gains, creating cross-frequency harmonization that strengthens spectral token operators when inserted into standard vision backbones.","pith_inferences":["The same shared-basis idea could be tested in other frequency-domain operators beyond Fourier mixers.","Coherent sampling geometry may prove important for realizing cross-frequency benefits in related architectures.","The structured separation of shared and specific components might help control parameter count while retaining adaptivity."],"forward_implications":["Higher reconstruction quality in accelerated MRI tasks compared to same-backbone baselines.","Improved segmentation accuracy on undersampled MRI data.","Better results in natural-image reconstruction using the same backbone.","Ablations confirm that dropping the shared-basis constraint weakens the observed benefit."],"fun_headline_variants":["CHASM aligns channel directions with shared eigenbasis across frequencies","Shared eigenbasis with frequency gains for spectral token mixers","Cross-frequency harmonization via CHASM spectral mixer","CHASM uses common channel basis for adaptive frequency responses"],"cache_read_input_tokens":64,"weakest_assumption_plain":"Enforcing a shared channel eigenbasis across frequencies supplies a useful inductive bias whose benefit is not merely an artifact of extra parameters or particular training setups.","fun_headline_variants_meta":{"raw":{"variants":["CHASM aligns channel directions with shared eigenbasis across frequencies","Shared eigenbasis with frequency gains for spectral token mixers","Cross-frequency harmonization via CHASM spectral mixer","CHASM uses common channel basis for adaptive frequency responses"]},"model":"grok-4.3","cost_usd":0.005344,"raw_usage":{"total_tokens":2480,"prompt_tokens":631,"num_sources_used":0,"completion_tokens":62,"cost_in_usd_ticks":53440500,"prompt_tokens_details":{"text_tokens":631,"audio_tokens":0,"image_tokens":0,"cached_tokens":64},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":1787,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":631,"tokens_out":62,"duration_ms":19044,"temperature":1.0,"reasoning_tokens":1787,"cache_read_input_tokens":64,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-05-15T04:35:05.204424+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"An experiment in which removing the shared-basis constraint leaves performance unchanged or randomizing coherent sampling geometry eliminates the reported gains.","supporting_citations":[],"review_version":1}