{"id":"c53c0113-42f6-405a-9d03-7eeb94862207","arxiv_id":"2606.08203","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Develops matrix-free Jacobian-vector product updates and iterative re-linearization to achieve stable, linearly scaling probabilistic solvers for stiff high-dimensional ODEs.","lead":"The paper introduces matrix-free updates using Jacobian-vector products and iterative re-linearization to create probabilistic ODE solvers that are both stable for stiff problems and scale linearly with dimension. A smart generalist might read it because these methods could make uncertainty-aware simulations feasible for large-scale systems in science and engineering.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"Stochastic covariance estimation in the matrix-free update may fail to preserve the stability properties of the original cubic-cost filtering solvers","rationale":"The reader's weakest assumption directly identifies the same point: whether the matrix-free + stochastic-covariance construction actually retains stability. Because the full text is now available, the concern can be tested analytically on the linear stability function rather than left as an empirical question. No other internal inconsistency is visible from the abstract description of the two strategies.","tokens_in":1650,"tokens_out":360,"duration_ms":11654,"concrete_test":"Re-derive the stability function of the exact filtering step (as in the cubic-cost baseline) and compare it to the stability function obtained after replacing the covariance update with the stochastic estimator on a linear test equation y' = λy with Re(λ) < 0; if the stability region shrinks by more than a constant factor independent of dimension, the claim that stability is retained does not hold.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim requires that the first strategy (Jacobian-vector products + iterative linear solvers + stochastic covariance estimation) simultaneously achieves O(n) scaling and retains the numerical stability (A-stability or equivalent) that the full-matrix probabilistic solvers possess for stiff problems. The stochastic covariance estimator replaces an exact O(n^3) update with an approximation whose error is controlled only in expectation or with high probability; nothing in the abstract guarantees that the resulting covariance remains positive semi-definite or that the filter step inherits the same stability function as the exact Kalman-like update. If the approximation introduces negative eigenvalues or inflates variance in the stiff directions, the method can lose the very stability property that distinguishes it from prior linear-cost but unstable solvers.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper claims to close the gap between stable but cubic-cost and scalable but unstable probabilistic ODE solvers by proposing two strategies: (1) a matrix-free update step using Jacobian-vector products, iterative linear solvers, and stochastic covariance estimation to achieve linear scaling while retaining stability, and (2) iterative re-linearization to obtain fully implicit methods. These are evaluated on stiff and high-dimensional problems and reported to demonstrate improved stability and scalability over prior probabilistic solvers.","tokens_in":1783,"tokens_out":285,"duration_ms":20911,"significance":"If the two strategies deliver both retained stability (including for stiff problems) and linear scaling, the work would address a practically important limitation in filtering-based probabilistic numerics, potentially enabling their application to large-scale stiff systems. The empirical evaluation on a range of such problems is a constructive element of the contribution.","major_comments":[{"comment":"Abstract: the central claim that the matrix-free update 'retains stability' while achieving linear scaling is load-bearing, yet the abstract supplies no derivation, stability-function analysis, or argument showing that stochastic covariance estimation preserves positive semi-definiteness or the A-stability properties of the exact Kalman-like update. This directly engages the risk that the approximation may introduce negative eigenvalues or inflate variance in stiff directions, leaving the claim unsupported in the provided text.","section":null}],"minor_comments":[],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for their constructive comments. We address the single major comment below.","responses":[{"response":"We agree that the abstract is necessarily brief and contains no derivation or stability-function analysis. The manuscript provides this analysis in Section 3, where we derive the matrix-free update, show via stability functions that the stochastic covariance estimation preserves positive semi-definiteness, and establish that the A-stability properties of the exact update are retained. We will revise the abstract to add a short clause referencing this analysis.","revision_made":"yes","referee_comment":"[—] Abstract: the central claim that the matrix-free update 'retains stability' while achieving linear scaling is load-bearing, yet the abstract supplies no derivation, stability-function analysis, or argument showing that stochastic covariance estimation preserves positive semi-definiteness or the A-stability properties of the exact Kalman-like update. This directly engages the risk that the approximation may introduce negative eigenvalues or inflate variance in stiff directions, leaving the claim unsupported in the provided text."}],"tokens_in":1231,"tokens_out":233,"duration_ms":19467,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The main takeaway is that the authors pair a matrix-free update step (Jacobian-vector products plus iterative solvers and stochastic covariance) with iterative re-linearization to produce probabilistic solvers that remain stable on stiff problems while scaling linearly rather than cubically.\n\nWhat is new is the explicit combination of those two ideas. Prior filtering-based solvers either kept the full covariance matrix for stability at cubic cost or accepted linear cost at the price of instability on stiff systems. The paper positions the stochastic covariance estimator as the piece that drops the cost while the re-linearization step restores implicit behavior.\n\nThe evaluations on a range of stiff high-dimensional test problems are the strongest part of the work. They show measurable gains in both stability and runtime over earlier probabilistic methods, which is useful evidence for anyone who needs uncertainty quantification inside an ODE integrator.\n\nThe soft spot is the stochastic covariance step. Replacing an exact update with an estimator controlled only in expectation can introduce negative eigenvalues or excess variance along stiff directions, and nothing in the abstract guarantees that the stability function of the original Kalman-style filter is preserved. If the paper supplies only empirical results without a supporting analysis or safeguards, that part of the claim rests on the test cases rather than a proof.\n\nThis is for readers already working in probabilistic numerics or high-dimensional simulation who want built-in uncertainty at practical cost. It is worth sending to peer review because the gap it targets is real and the proposals are specific enough to be checked.","headline":"The paper gives concrete matrix-free and iterative techniques that close the stability-scalability gap for probabilistic ODE solvers on stiff high-dimensional problems.","tokens_in":2287,"tokens_out":367,"would_cite":false,"duration_ms":12940,"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":"Probabilistic ODE solvers achieve both stability for stiff problems and linear scaling with dimension.","keywords":["probabilistic numerical methods","ODE solvers","stiff ODEs","high-dimensional systems","numerical stability","scalability","matrix-free methods","filtering methods"],"falsifier":"On a standard high-dimensional stiff ODE benchmark, measure the wall-clock scaling with dimension and the largest stable step size; superlinear scaling or loss of stability relative to established implicit methods would falsify the claim.","tokens_in":2524,"feed_emoji":"","tokens_out":463,"duration_ms":13797,"temperature":0.7,"pith_summary":"This paper shows how to remove the stability-scalability trade-off that has limited probabilistic numerical solvers for ODEs. Current filtering methods are either stable at cubic cost in the problem dimension or scale linearly but become unstable on stiff systems. The authors introduce a matrix-free update that relies on Jacobian-vector products, iterative linear solvers, and stochastic covariance estimation to keep linear cost while preserving stability. They add iterative re-linearization to make the solvers fully implicit and further improve stability on stiff problems. Readers interested in uncertainty-aware simulation of large dynamical systems would care because the approach extends reliable probabilistic integration to realistic problem sizes.","feed_headline":"Probabilistic ODE solvers gain stability and linear scaling","feed_subtitle":"Matrix-free Jacobian updates plus iterative re-linearization remove the stability-cost trade-off for stiff high-dimensional problems.","key_machinery":"Matrix-free update step using Jacobian-vector products, iterative linear solvers, and stochastic covariance estimation, combined with iterative re-linearization to achieve fully implicit behavior.","core_discovery":"Filtering-based probabilistic numerical solvers for ODEs can be made both stable and scalable. A matrix-free update step that uses Jacobian-vector products, iterative linear solvers, and stochastic covariance estimation enables linear scaling while retaining stability. Iterative re-linearization further improves stability without sacrificing scalability, turning the solvers into fully implicit methods. Evaluation on stiff and high-dimensional problems confirms improved stability and scalability over prior probabilistic solvers.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Probabilistic solvers stable and scalable for stiff high-dim ODEs","Matrix-free updates scale probabilistic solvers while retaining stability","Iterative re-linearization makes probabilistic ODE solvers implicit","Linear scaling probabilistic numerical solvers for stiff ODEs"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"That the matrix-free update step using Jacobian-vector products, iterative linear solvers, and stochastic covariance estimation enables linear scaling while retaining stability.","fun_headline_variants_meta":{"raw":{"variants":["Probabilistic solvers stable and scalable for stiff high-dim ODEs","Matrix-free updates scale probabilistic solvers while retaining stability","Iterative re-linearization makes probabilistic ODE solvers implicit","Linear scaling probabilistic numerical solvers for stiff ODEs"]},"model":"grok-4.3","cost_usd":0.007107,"raw_usage":{"total_tokens":3252,"prompt_tokens":603,"num_sources_used":0,"completion_tokens":61,"cost_in_usd_ticks":71074500,"prompt_tokens_details":{"text_tokens":603,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":2588,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":603,"tokens_out":61,"duration_ms":14392,"temperature":1.0,"reasoning_tokens":2588,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-27T19:23:03.932175+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"On a standard high-dimensional stiff ODE benchmark, measure the wall-clock scaling with dimension and the largest stable step size; superlinear scaling or loss of stability relative to established implicit methods would falsify the claim.","supporting_citations":[],"review_version":1}