{"id":"dbc66338-8915-4044-8448-376d85cdf656","arxiv_id":"2606.06982","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"K-recursion plus automatic differentiation yields a topology-agnostic optimizer for end-to-end mutual information in linear Gaussian wireless DAGs that recovers water-filling optima where known.","lead":"The paper introduces a K-recursion to compute all node-pair covariances in linear Gaussian DAGs, then uses reverse-mode automatic differentiation to obtain exact gradients for end-to-end mutual information maximization under global power constraints. A smart generalist might read it to see how differentiable programming can automate network design across MIMO, relays, and RIS without custom per-topology derivations.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"K-recursion correctness for non-adjacent cross-covariances in branching/merging DAGs rests only on numerical matches to known optima","rationale":"The reader's weakest assumption is exactly the load-bearing step. The numerical evidence in the four test topologies is consistent with correctness but does not constitute an independent check for arbitrary DAGs; a direct covariance comparison on one non-trivial topology would settle the issue without requiring new experiments.","tokens_in":1803,"tokens_out":336,"duration_ms":12523,"concrete_test":"For the diamond DAG in §4, compute the exact joint covariance matrix of source and sink analytically from the linear model; compare entry-by-entry with the matrix returned by the K-recursion at the same parameter values. Agreement to machine precision on all entries (including non-adjacent cross terms) confirms the recursion; discrepancy >1e-10 falsifies it.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The strongest claim requires that the K-recursion (topological propagation of all node-pair covariances) produces exactly the joint statistics needed for the log-det end-to-end MI expression to equal the true I(source; sink) for any linear Gaussian DAG. The paper demonstrates agreement with water-filling on single-link MIMO and reports MI gains on diamond, two-hop AF, and 11-node multi-layer cases, but supplies no independent derivation or exhaustive check that cross terms between non-adjacent nodes are assembled without omission or duplication when paths merge or branch. Because the gradient is obtained by differentiating this recursion, any systematic error in the covariance assembly would propagate directly into the reported optima.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript proposes a topology-agnostic differentiable framework for end-to-end mutual information optimization over linear Gaussian DAGs. Its core is a K-recursion that propagates all node-pair covariances (including non-adjacent cross terms) in topological order to yield a closed-form log-determinant MI expression; reverse-mode automatic differentiation then supplies exact Wirtinger gradients, which are used with projected gradient ascent under global constraints such as total power. The same implementation is applied to single-link MIMO, diamond DAG, two-hop AF relay, input-covariance shaping, and an 11-node multi-layer network, with the claim that it recovers water-filling optima where available and yields improvements otherwise.","tokens_in":1937,"tokens_out":452,"duration_ms":14183,"significance":"If the K-recursion is shown to be correct, the framework would provide a practical, general-purpose tool for MI maximization in wireless networks with branching/merging topologies where closed-form per-topology gradients are unavailable. The use of automatic differentiation on an analytic recursion is a methodological strength that avoids manual gradient derivations.","major_comments":[{"comment":"Abstract: the central claim that 'the same implementation reaches the classical water-filling optimum in the settings where it is available and yields MI improvements' is unsupported by any numerical results, tables, error metrics, or comparison data in the manuscript, so the accuracy of the K-recursion and the reported gains cannot be assessed.","section":"Abstract"},{"comment":"K-recursion (core ingredient described in Abstract): correctness for non-adjacent cross-covariances in branching and merging paths is load-bearing for the end-to-end log-det MI to equal the true I(source;sink); the manuscript supplies only agreement with water-filling on single-link cases and does not provide an independent derivation or exhaustive verification that all required covariances are assembled without omission or duplication when paths merge or branch.","section":"Abstract"}],"minor_comments":[],"recommendation":"major_revision","confidential_remarks":"The absence of any numerical verification or comparison data despite explicit performance claims in the abstract is a serious completeness issue for a methods paper in information theory."},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the careful reading and constructive comments on our manuscript. We address each major comment below, pointing to the existing support in the paper and indicating where revisions will strengthen clarity and verification.","responses":[{"response":"Sections IV and V present the numerical results: single-link MIMO recovers the exact water-filling MI; the diamond DAG, two-hop AF relay, and 11-node network report optimized MI values with comparisons to baseline schemes showing gains. To address the concern about explicit support, we will add a summary table with quantitative MI values, error metrics, and direct comparisons in the revised manuscript.","revision_made":"yes","referee_comment":"[Abstract] Abstract: the central claim that 'the same implementation reaches the classical water-filling optimum in the settings where it is available and yields MI improvements' is unsupported by any numerical results, tables, error metrics, or comparison data in the manuscript, so the accuracy of the K-recursion and the reported gains cannot be assessed."},{"response":"Section III derives the K-recursion via induction on topological order, with explicit rules for propagating all pairwise covariances (including non-adjacent cross terms) to correctly handle branching and merging without omission or duplication. Single-link agreement is the base case; the multi-node experiments provide verification on branching topologies. We will add an appendix with a small branching-DAG example showing manual covariance computation versus the recursion output.","revision_made":"partial","referee_comment":"[Abstract] K-recursion (core ingredient described in Abstract): correctness for non-adjacent cross-covariances in branching and merging paths is load-bearing for the end-to-end log-det MI to equal the true I(source;sink); the manuscript supplies only agreement with water-filling on single-link cases and does not provide an independent derivation or exhaustive verification that all required covariances are assembled without omission or duplication when paths merge or branch."}],"tokens_in":1490,"tokens_out":421,"duration_ms":29384,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The paper's main contribution is a forward K-recursion that walks any linear Gaussian DAG in topological order and assembles every required node-pair covariance, including the non-adjacent cross terms that appear when paths branch or merge. Reverse-mode Wirtinger differentiation then produces the gradient of the log-det mutual information with respect to all controllable parameters in one backward pass, after which projected gradient ascent handles the global power constraint. The same implementation is run on single-link MIMO, a diamond network, a two-hop AF relay, and an 11-node multi-layer graph.\n\nWhat the work does well is remove the need to derive a fresh gradient expression for each new topology. The numerical results show the code recovers the classical water-filling solution on the single-link case and produces measurable MI gains on the other three topologies. Those matches supply concrete evidence that the covariance assembly is accurate enough for the tested graphs.\n\nThe remaining soft spot is the lack of an independent algebraic check that the recursion produces exactly the joint covariance matrix needed for the end-to-end MI formula on arbitrary DAGs. The paper rests on numerical agreement rather than a separate proof or exhaustive enumeration of cross terms. A systematic off-by-one error in merging paths would propagate straight into the gradients, though the water-filling match lowers the practical risk.\n\nThis is useful for communications researchers who routinely optimize MIMO, relay, or RIS parameters under total-power budgets and want one reusable optimizer instead of case-by-case derivations. A reader working on network-level design problems would get immediate value from the framework.\n\nThe paper deserves a serious referee because the core mechanism is new in its generality and the experiments are concrete. I would send it to review.","headline":"The K-recursion with automatic differentiation supplies a single code path for MI optimization across linear Gaussian DAGs, and the reported matches to water-filling give reasonable evidence it works.","tokens_in":2446,"tokens_out":422,"would_cite":false,"duration_ms":13630,"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":"K-recursion assembles node-pair covariances for any linear Gaussian DAG so automatic differentiation supplies exact gradients for mutual-information maximization under global power limits.","keywords":["mutual information optimization","linear Gaussian DAGs","K-recursion","automatic differentiation","wireless networks","MIMO precoding","amplify-and-forward relays"],"falsifier":"Run the framework on a single-link MIMO channel whose water-filling mutual information is known in closed form and verify whether both the attained value and the computed gradients match the classical solution to machine precision.","tokens_in":2694,"feed_emoji":"📡","tokens_out":769,"duration_ms":13740,"temperature":0.7,"pith_summary":"The paper develops a single differentiable framework that optimizes end-to-end mutual information over linear Gaussian directed acyclic graphs representing wireless networks. Its central device is a K-recursion that propagates every required covariance, including non-adjacent cross terms, in topological order. The resulting covariances enter a closed-form log-determinant mutual-information expression that admits exact Wirtinger gradients via reverse-mode automatic differentiation. Projected gradient ascent then maximizes this expression subject to a total transmit-power budget. The same code reaches the classical water-filling solution on single-link MIMO channels and produces measurable gains on diamond, two-hop relay, and eleven-node multi-layer topologies without any topology-specific gradient derivation.","feed_headline":"K-recursion supplies exact gradients for MI maximization in any Gaussian DAG","feed_subtitle":"One implementation reaches water-filling optima and improves diamond and relay topologies under total power constraints.","key_machinery":"The K-recursion, which computes every node-pair covariance matrix by forward propagation in topological order so that the subsequent log-determinant expression equals the true end-to-end mutual information.","core_discovery":"The K-recursion analytically propagates all node-pair covariances along the DAG in topological order, including non-adjacent cross-covariances required for branching and merging paths. These covariances yield a closed-form log-determinant expression for the end-to-end mutual information as a smooth function of the controllable factors. Complex-valued reverse-mode automatic differentiation on the K-recursion returns the exact Wirtinger gradient at every controllable factor in a single backward sweep. Projected gradient ascent then maximizes the mutual information under global constraints such as a total transmit power budget, and the identical implementation applies without modification to si","pith_inferences":["The covariance recursion could be reused for other differentiable network objectives such as secrecy rates or estimation error provided the objective remains a smooth function of the covariances.","If the underlying graph contains cycles, replacing the topological traversal with a fixed-point iteration on the covariance equations would preserve the automatic-differentiation step.","The framework supplies a concrete testbed for checking whether learned precoders or relay gains generalize across randomly generated DAG topologies of fixed depth."],"forward_implications":["The identical code reaches the classical water-filling optimum on single-link MIMO channels.","The same implementation yields mutual-information gains on diamond DAGs and two-hop amplify-and-forward relays.","Input-covariance shaping is optimized inside the same topology-agnostic procedure.","The method extends without change to an eleven-node, five-layer Gaussian network for which no closed-form gradient exists."],"fun_headline_variants":["K-recursion yields exact MI gradients in Gaussian DAGs","Automatic differentiation on K-recursion for MI in any DAG","Single K-recursion handles MI optimization for all Gaussian topologies","K-recursion and AD maximize MI without topology-specific formulas"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The K-recursion must correctly assemble every required node-pair covariance, including non-adjacent cross terms, so that the log-determinant expression equals the true end-to-end mutual information.","fun_headline_variants_meta":{"raw":{"variants":["K-recursion yields exact MI gradients in Gaussian DAGs","Automatic differentiation on K-recursion for MI in any DAG","Single K-recursion handles MI optimization for all Gaussian topologies","K-recursion and AD maximize MI without topology-specific formulas"]},"model":"grok-4.3","cost_usd":0.007825,"raw_usage":{"total_tokens":3630,"prompt_tokens":785,"num_sources_used":0,"completion_tokens":65,"cost_in_usd_ticks":78249500,"prompt_tokens_details":{"text_tokens":785,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":2780,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":785,"tokens_out":65,"duration_ms":13902,"temperature":1.0,"reasoning_tokens":2780,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-27T21:09:05.275452+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"Run the framework on a single-link MIMO channel whose water-filling mutual information is known in closed form and verify whether both the attained value and the computed gradients match the classical solution to machine precision.","supporting_citations":[],"review_version":1}