{"id":"ce79427e-d964-4e1c-9872-1c926e0786d5","arxiv_id":"2508.01013","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":5.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":2,"one_line_summary":"A proximity-based acquisition function for multi-fidelity Bayesian optimization controls how often the expensive model is queried, while a multi-fidelity UCB variant is benchmarked on chemical kinetics problems.","lead":"This paper tests a new 'proximity' rule for deciding when to run expensive, high-fidelity simulations during optimization, rather than relying only on cheap low-fidelity models. It also combines multi-fidelity Gaussian process surrogates with upper confidence bound strategies, and applies them to chemical reaction models.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proximity-based fidelity selection lacks a stated criterion for when proximity justifies skipping high-fidelity queries, so the 'consistent control' claim may rest on an uncalibrated scale rather than a principled rule.","rationale":"The reader correctly identified the proximity proxy as the weakest assumption. My stress-test adds that the proxy is not just an empirical shortcut but a potentially uncalibrated scale: the abstract does not state what 'proximity' means mathematically, so the reported control over high-fidelity usage could be a tuning artifact. This concern cannot be resolved from the abstract alone; the paper remains unverdictable pending full-text and code. I therefore keep the reader's UNVERDICTED verdict. The agreement is partial because I focus on the missing definition and the need for a trivial baseline comparison, while the reader framed the issue in terms of 'reliable indicator' and premature stopping. Both formulations point to the same core step, but my proposed concrete test is more operational.","tokens_in":734,"tokens_out":1697,"duration_ms":24648,"concrete_test":"Read the full text to extract the exact definition of the proximity measure and the acquisition rule (likely in Section 3 or Algorithm 1). Then implement the method and run the same benchmarks (including the chemical kinetic examples) with fixed GP priors, varying only the proximity parameter. Record the Pareto frontier of cumulative high-fidelity evaluations versus optimization regret. If any point on this frontier does not strictly dominate the always-low-fidelity baseline in convergence, and if the 'consistent control' disappears when the objective is globally scaled, the claim is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's central claim is that the proximity-based acquisition function 'delivers consistent control over high-fidelity usage while maintaining convergence efficiency.' The implied mechanism is that the distance between the current GP posterior and a reference value (e.g., the last high-fidelity observation) indicates when additional high-fidelity information is unnecessary. However, no definition of 'proximity' is given: it is not specified whether this is Euclidean distance in input space, a function of predictive variance, a threshold on posterior mean differences, or something else. Without this definition, the tunable 'proximity' parameter could simply rescale the acquisition function's preference for low fidelity, yielding an apparently controllable trade-off that is not guaranteed to preserve convergence. In particular, if the proximity metric is not invariant to the scale of the objective or the GP hyperparameters, the same parameter value may produce very different high-fidelity budgets across the benchmark problems, and the reported consistency may be an artifact of problem-specific tuning. The abstract also gives no comparison of the proximity method to a trivial baseline that always queries the low-fidelity model; 'maintaining convergence efficiency' is only meaningful relative to such a baseline. Because the full text is unavailable, this is an unresolved risk in the central argument rather than a demonstrated flaw, but it is the load-bearing assumption on which the headline claim depends.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes two tunable multi-fidelity Bayesian optimization frameworks: a proximity-based acquisition function intended to replace the need for separate acquisition functions at each fidelity level, and an extension of multi-fidelity Upper Confidence Bound (UCB) strategies to multi-fidelity Gaussian processes. The authors report benchmarking these approaches against other multi-fidelity acquisition strategies, with particular emphasis on 'consistent control over high-fidelity usage while maintaining convergence efficiency.' The examples include multi-fidelity chemical kinetic models, both homogeneous and heterogeneous.","tokens_in":956,"tokens_out":2143,"duration_ms":25303,"significance":"If the claims hold, the proximity-based acquisition function would be a practical simplification of multi-fidelity Bayesian optimization, removing the need to design separate acquisition functions per fidelity level. The extension of UCB to multi-fidelity GPs also broadens the toolkit for such problems. However, the abstract alone provides no quantitative results, derivations, or error bars, so the empirical significance cannot be assessed from the submitted text. The framing of the problem and the proposed unification are conceptually interesting, but the evidence needed to support the central claim is not visible in the abstract.","major_comments":[{"comment":"The abstract states that the proximity-based acquisition function 'delivers consistent control over high-fidelity usage while maintaining convergence efficiency,' but it does not define the proximity metric. Without specifying whether proximity is measured in input space, function-value space, or via posterior uncertainty, and without stating how the proximity threshold is normalized across problems with different objective scales, the central claim is not testable from the provided text.","section":"Abstract"},{"comment":"The abstract reports benchmark outcomes without presenting any quantitative results, such as regret curves, high-fidelity budgets, or error bars, and without a comparison to a trivial baseline such as always querying the low-fidelity model. The claim of 'maintaining convergence efficiency' is only meaningful relative to such a baseline, and its absence leaves the central empirical claim unsupported.","section":"Abstract"},{"comment":"The proximity-based fidelity selection rests on the assumption that the nearness of the current posterior to a reference value (for example, a previous high-fidelity observation) indicates when additional high-fidelity information is unnecessary. The abstract provides no derivation or justification for this proxy, which is load-bearing because an unreliable proxy could either waste high-fidelity queries or cause premature convergence.","section":"Abstract"}],"minor_comments":[{"comment":"The title 'On Some Tunable Multi-fidelity Bayesian Optimization Frameworks' is vague; a more specific title mentioning the proximity-based acquisition function and the multi-fidelity UCB extension would better convey the contributions.","section":"Title"},{"comment":"The phrase 'rather than the standard GPs typically used' would benefit from clarification: it is unclear which particular GP model is considered standard and how the multi-fidelity GP differs from it in this context.","section":"Abstract"},{"comment":"The abstract mentions 'hyperparameter tunability' as a benchmark dimension, but it does not indicate how tunability is measured or compared across methods; a brief clarification would help readers interpret the claimed advantage.","section":"Abstract"}],"recommendation":"uncertain","confidential_remarks":"This review is based solely on the abstract, as the full text is not available. The central claims are empirical and cannot be verified without the mathematical definitions of proximity and the experimental details. I recommend obtaining the full manuscript before making an editorial decision, and I would be willing to review the full text if it is provided."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a plausible, modestly novel empirical paper on multi-fidelity BO, but the abstract alone cannot support the headline claim about \"consistent control.\" Worth sending to a referee if the full text defines its terms.\n\nThe actual new bits: a proximity-based acquisition rule that removes the need for separate acquisition functions per fidelity, and a UCB version built on multi-fidelity GPs rather than standard ones. Benchmarking on chemical kinetics, including a dynamic catalysis problem, is a reasonable choice, and the paper frames a practical pain point: one knob for fidelity control instead of a menu of acquisition functions.\n\nWhat it does well: the idea is simple to state, and the paper acknowledges multiple comparators, including fidelity-weighted strategies. If the code and data are released, this could be a useful reference for practitioners.\n\nWhere I'm cautious: the central claim is empirical, and the abstract gives zero numbers, error bars, or a formal definition of \"proximity.\" The stress-test note hits the right spot. Without a definition of when proximity justifies skipping a high-fidelity query, \"consistent control\" could just mean the parameter rescales low-fidelity preference, and the consistency could be a tuning artifact. Also, \"maintaining convergence efficiency\" needs a baseline: they mention comparing with other multi-fidelity methods, but a low-fidelity-only baseline would be the minimal sanity check. These are unresolved risks, not demonstrated flaws; the full text might define everything carefully. I can't see the math, so I'm unverdictive on soundness, but not suspicious.\n\nThe citation pattern is invisible from the abstract; no red flags there.\n\nThis is a serious, workmanlike contribution for researchers running multi-fidelity BO with tight budgets. It deserves a real peer review, not a desk reject, because the algorithm design is concrete and the benchmark is domain-relevant. A referee should pin down the proximity metric and the sensitivity to the tuning parameter before accepting the consistency claim. I wouldn't cite it until I've seen the full results, but I'd read the next version.","headline":"A plausible tunable multi-fidelity BO paper whose headline claim rests on an undefined proximity criterion; send to review but ask for definitions.","tokens_in":1509,"tokens_out":2137,"would_cite":false,"duration_ms":25796,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A proximity-based acquisition function is claimed to control high-fidelity usage in multi-fidelity Bayesian optimization with one tunable rule.","keywords":["multi-fidelity Bayesian optimization","Gaussian process surrogates","proximity-based acquisition","Upper Confidence Bound","chemical kinetics","ammonia production catalysis","fidelity selection","hyperparameter tuning"],"falsifier":"On the paper's own benchmarks, vary the proximity threshold from very tight to very loose and record both the number of high-fidelity evaluations and the final objective error. If the high-fidelity count does not respond monotonically to the knob, or if moderate settings produce worse optima than low-fidelity-only optimization, the claimed consistent control over high-fidelity usage fails.","tokens_in":547,"feed_emoji":"🎯","tokens_out":4865,"duration_ms":55600,"temperature":0.7,"pith_summary":"This paper is trying to establish that a single acquisition function, built on proximity between the Gaussian process surrogate and a reference value, can take over fidelity selection in multi-fidelity Bayesian optimization. The pay-off of the claim is practical: instead of maintaining a separate acquisition function for each fidelity level, the user tunes one parameter that trades high-fidelity evaluations against convergence speed. The paper also argues that Upper Confidence Bound strategies work with multi-fidelity Gaussian processes, not only with standard GPs. Benchmark results on representative tasks, including chemical kinetic models for ammonia production, are presented as evidence that the proximity-based approach gives consistent control over high-fidelity usage without sacrificing convergence efficiency. If the claim holds, tuning an expensive simulation campaign becomes a one-knob problem rather than a per-fidelity design exercise.","feed_headline":"Proximity rule tunes how often costly simulators run","feed_subtitle":"Blending all fidelities into one tunable rule cuts dependence on expensive high-fidelity runs.","key_machinery":"The load-bearing object is the proximity-based multi-fidelity acquisition function: a single scalar rule that scores a candidate point by the proximity of the current Gaussian process posterior to a reference value, and uses that score both to select the next evaluation and to choose the fidelity level. Its role in the argument is to replace the usual stack of acquisition functions, one per fidelity. The second piece of machinery is the pairing of multi-fidelity UCB acquisition with multi-fidelity Gaussian processes, which lets exploration-exploitation trade-offs be expressed at every fidelity rather than only at the highest. Together the two pieces are what make high-fidelity usage a tunable, continuously adjustable quantity.","core_discovery":"On the paper's own terms, the central discovery is that fidelity selection can be driven by how close the current Gaussian process posterior stands to a reference value rather than by a portfolio of fidelity-specific acquisition functions. The proximity signal is used both to propose the next query and to decide when a low-fidelity evaluation suffices, so that high-fidelity calls are spent only when the surrogate is far from the reference. Implemented with multi-fidelity GPs, the same proximity rule extends to Upper Confidence Bound acquisition, yielding a tunable strategy whose high-fidelity usage remains controllable across tasks. The illustrative applications to homogeneous and heterogeneous chemical kinetics, including a dynamic catalysis model for ammonia production, support the claim that this control survives in practically relevant simulators.","pith_inferences":["A natural extension would be to replace the fixed reference value with an adaptive target, such as the current best posterior mean, which could make the proximity rule self-tuning rather than dependent on a user-supplied reference.","If the proximity criterion is a reliable need-for-information signal, it could be combined with information-theoretic acquisition functions to yield a multi-fidelity entropy search with the same one-knob property.","A direct test outside the paper's examples: on a fixed benchmark, sweep the proximity threshold and record high-fidelity count versus final error; the claim predicts a monotone trade-off curve that remains usable across settings."],"forward_implications":["One knob replaces per-fidelity acquisition design: a practitioner tunes proximity sensitivity instead of writing separate acquisition functions for cheap and expensive simulators.","Multi-fidelity UCB becomes usable with multi-fidelity GPs, so exploration-exploitation balance can be set once and applied across fidelities.","Expensive simulation campaigns, such as chemical kinetic models for ammonia catalysis, can budget high-fidelity evaluations explicitly while preserving convergence.","The same acquisition rule should extend to other multi-fidelity GP formulations, since the proximity signal is defined on the posterior rather than on a specific covariance structure."],"supporting_citations":[],"fun_headline_variants":["Proximity rule sets when simulators run at high fidelity","One rule controls fidelity use in Bayesian optimization","Proximity-based acquisition tunes multi-fidelity optimization","Tunable multi-fidelity optimization without extra acquisition functions","Proximity signal decides when costly simulators are needed"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The method assumes that the proximity between the surrogate's prediction and a reference value reliably indicates when low-fidelity evaluations are sufficient, so that skipping high-fidelity calls at close points does not silently degrade the optimum.","fun_headline_variants_meta":{"raw":{"variants":["Proximity rule sets when simulators run at high fidelity","One rule controls fidelity use in Bayesian optimization","Proximity-based acquisition tunes multi-fidelity optimization","Tunable multi-fidelity optimization without extra acquisition functions","Proximity signal decides when costly simulators are needed"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000501,"raw_usage":{"total_tokens":2404,"prompt_tokens":850,"completion_tokens":1554,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":466,"completion_tokens_details":{"reasoning_tokens":1476}},"tokens_in":466,"tokens_out":1554,"duration_ms":10796,"temperature":1.0,"reasoning_tokens":1476,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T05:52:03.261555+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"On the paper's own benchmarks, vary the proximity threshold from very tight to very loose and record both the number of high-fidelity evaluations and the final objective error. If the high-fidelity count does not respond monotonically to the knob, or if moderate settings produce worse optima than low-fidelity-only optimization, the claimed consistent control over high-fidelity usage fails.","supporting_citations":[],"review_version":1}