{"id":"c14354b7-9bd6-4159-a2fa-cec52562aabb","arxiv_id":"2502.03167","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"An 8-node oscillator-based Ising machine built from analog parts solves small max-cut instances, but evidence is anecdotal and per-problem tuned.","lead":"The authors built an eight-oscillator analog computer that maps small max-cut problems onto phase relations between coupled oscillators, and report that it returns correct partitions on simple test graphs. The work is a proof-of-concept: it reports no success rates, raw data are on request, and coupling strengths are tuned per problem.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Claim of 'consistently accurate solutions' is untestable: coupling strength is tuned per problem and raw data are not reported.","rationale":"The paper is best read as a construction note plus a qualitative demonstration. The strongest claim that the machine 'consistently produced accurate solutions' requires evidence that, for a fixed parameter setting, a large fraction of repeated runs yields the true max-cut. Section 10 instead reports that the coupling strength must be chosen individually per problem and that some problems fail without SHIL, and it gives no run counts or cut values. This means the headline result could be an artifact of per-problem search. I do not think this requires rejecting the paper; the hardware exists and the qualitative observations are useful. But the claim should be downgraded or conditioned on a public dataset and a pre-registered tuning protocol. Because the reader's CONDITIONAL verdict already captures this need, my read does not change the verdict.","tokens_in":9457,"tokens_out":5693,"duration_ms":53432,"concrete_test":"Fix a single protocol: for each reported graph (house, rectangle, toffee, crisscross12_78bar, hourglass, and any others), set all coupling weights to c=0.1, SHIL to the stated rectangular 2.5 V/6.8 kHz signal, and run 100 trials per graph with random initial phases; record the phase-derived partition and compare to the true max-cut computed by brute force (n<=8). If the fraction of trials reaching the true max-cut is not high (and stable across a second fixed c, e.g., 0.2), the 'consistently produced accurate solutions' claim fails and the per-problem tuning is exposed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in Section 10 — that the eight-oscillator machine 'consistently produced accurate solutions' when coupling weights follow the graph and a SHIL signal is applied — is not falsifiable from the paper as written. The authors state that 'the actual value c has to be chosen individually for every problem to achieve best results' and that SHIL parameters (waveform, amplitude, frequency) also matter, with no protocol fixing them in advance. If the experimenter tunes these parameters until the readout matches the known optimum, the reported success is an artifact of search rather than evidence of the dynamics solving max-cut. No quantitative results are given: no graph list, no cut sizes, no number of runs, no success rate; 'raw data and results ... will be provided by the authors upon request.' The paper itself acknowledges that closed odd loops 'partitioning into different configurations upon repeated runs,' so the machine does not reliably reach a global optimum for all tested topologies. Thus the strongest claim rests on an unquantified, per-problem parameter selection with no observed repeatability statistics.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper describes an experimental oscillator-based Ising machine built from off-the-shelf analog components, starting with a four-oscillator prototype and expanding to an eight-oscillator system. The hardware uses phase-shift oscillators coupled through a matrix of digital potentiometers, with second-harmonic injection locking (SHIL) to binarize oscillator phases into 0 or π. The authors map max-cut problems onto the coupling network, with antiphase synchronization representing the partition, and report that the eight-oscillator machine 'consistently produced accurate solutions' for the small max-cut instances tested. The paper also discusses synchronization theory, the Kuramoto model, and a list of open research questions.","tokens_in":9606,"tokens_out":3153,"duration_ms":30957,"significance":"If the central experimental claim were properly supported, this work would provide a low-cost, accessible hardware platform for exploring oscillator-based Ising machines, with useful implementation details such as oscillator schematics, phase-detector circuits, and a description of the hybrid analog-digital control interface. The paper is honest about several limitations, including the need for per-problem coupling-strength tuning and the observation of run-to-run variability on odd-loop graphs. However, as submitted, the experimental evidence is qualitative and underreported: no quantitative results are given, no repeated-run statistics are provided, and the raw data are only available upon request. The claimed demonstration of solving max-cut problems is therefore not verifiable from the manuscript alone, which is a serious shortcoming for a paper whose contribution is an experimental demonstration.","major_comments":[{"comment":"The central experimental claim is not supported by quantitative evidence. The section states that the configuration 'consistently produced accurate solutions' but provides no list of test graphs, no cut sizes or energies, no number of runs, no success rates, and no error bars. The term 'consistently' is undefined and cannot be checked by a reader. The statement that 'raw data and results ... will be provided by the authors upon request' does not meet the standard of a reproducible experimental report, and it makes the main finding unfalsifiable from the paper as written.","section":"Section 10"},{"comment":"The coupling strength c is chosen individually per problem, as stated: 'the actual value c has to be chosen individually for every problem to achieve best results.' Since the correct max-cut is evidently known when this choice is made, the procedure is vulnerable to circularity: the experimenter can tune c until the readout matches the known optimum. The paper gives no independent protocol for setting c (for example, a fixed rule based on graph statistics) and does not quantify how sensitive the outcome is to c. This undermines the claim that the machine solves max-cut problems rather than being fitted to the known answers.","section":"Section 10"},{"comment":"The paper acknowledges run-to-run variability, stating that closed odd loops lead to systems 'partitioning into different configurations upon repeated runs,' yet it reports no statistics over repeated measurements. It is therefore unclear whether the 'accurate solutions' represent typical behavior, best-case behavior, or manually selected runs. Without repeated-run statistics, the claim of 'consistently produced accurate solutions' is not falsifiable. This is a load-bearing issue because the existence of incorrect configurations shows that the dynamics do not always reach the ground state, and the paper does not quantify the frequency or conditions of such failures.","section":"Section 10"}],"minor_comments":[{"comment":"The oscillator resonance frequency is stated as about 3.8 kHz, but the SHIL signal in Section 10 is reported at 6.8 kHz. Since the second harmonic of a 3.8 kHz signal would be 7.6 kHz, the relationship between the oscillator frequency and the SHIL frequency should be clarified or reconciled.","section":"Section 9"},{"comment":"The mapping between edge weights and coupling coefficients is given as J = -μij, but the sign convention and its connection to the Ising Hamiltonian in Eq. (1) are not explained. A brief derivation or reference would help the reader understand how the physical coupling realizes the max-cut objective.","section":"Section 8"},{"comment":"The number of coupling weights is described as 'n × n − n,' which is mathematically correct but would be clearer as n(n-1). Additionally, the text should clarify whether the potentiometer matrix includes the diagonal (with zero values) as implied by Figure 6.","section":"Section 9"},{"comment":"The discussion of complexity classes (P, NP, NP-hard) is textbook material that is not specific to the present contribution and could be condensed substantially to keep the paper focused on the experimental results.","section":"Section 3"},{"comment":"The references include non-archival or informal sources (for example, Marandi [2019] is a YouTube video and the Stanford G-set is cited with an access date). For a journal submission, please ensure that all key claims about prior Ising-machine work are supported by peer-reviewed references.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The manuscript reads like an extended abstract or workshop paper: the hardware description is valuable, but the experimental section lacks the quantitative data that would make the central claim credible. The authors should be asked to provide full experimental results, including graph instances, cut values, parameter settings, and repeated-run statistics, before the paper can be considered for publication. If the raw data are truly not available in reproducible form, the claim of solving max-cut problems should be substantially weakened."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is what it says on the tin: a build report for an eight-oscillator analog Ising machine using off-the-shelf components, with a handful of max-cut experiments. The genuinely new bit is the concrete implementation—phase-shift oscillators, digital-potentiometer coupling matrix, SHIL injection, simple phase detectors—and the qualitative observations about which graphs cause trouble. That is a legitimate contribution, especially for researchers who want to reproduce or extend a low-cost OIM in a teaching lab. The authors also deserve credit for being transparent about the hard parts: they say outright that the coupling strength c has to be tuned per problem, that odd loops partition differently on repeated runs, and that G-set benchmarks have not been done. That honesty is the paper's best feature.\n\nThe soft spot is exactly the one the stress-test flags. Section 10 says the machine 'consistently produced accurate solutions,' but no raw numbers appear anywhere: no graph list, no cut sizes, no number of runs, no success rate. Data is 'provided upon request.' Combined with per-problem tuning of c and SHIL parameters, the central claim is not falsifiable from the text. If an experimenter tunes until the readout matches the known optimum, the demonstration partially reduces to parameter fitting. The paper even admits the machine does not reliably converge for odd-loop topologies, so 'consistently' is doing too much work. This is not a fatal flaw for a preliminary engineering note, but it is a load-bearing weakness: as written, the evidence is anecdotal.\n\nI also want to note a minor issue: the abstract and intro overstate the promise of Ising machines relative to what is shown. The 'groundbreaking improvements' language belongs in a grant proposal, not an experimental report. The tutorial sections (NP-hardness, Kuramoto model) are fine but could be compressed.\n\nWho is this for? Anyone considering building a similar analog OIM, or teaching unconventional computing. It does not advance the theory or give a benchmarked solver. A serious referee should engage with it, but the decision should be conditional on the authors either providing the raw data and run statistics or explicitly reframing the paper as a hardware description with preliminary observations rather than a performance claim.\n\nMy recommendation: send it to review, but make clear the evidence needs to be quantified before publication.","headline":"An honest but under-evidenced report on a small analog oscillator Ising machine; worth refereeing as a proof-of-concept, but the central 'consistently accurate' claim needs quantitative backing.","tokens_in":10167,"tokens_out":1077,"would_cite":false,"duration_ms":12535,"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":"An eight-oscillator analog Ising machine solves small max-cut graphs by phase locking.","keywords":["Ising machine","oscillator network","max-cut","antiphase synchronization","SHIL","analog computing","phase-shift oscillator","combinatorial optimization"],"falsifier":"Run the machine on an 8-node max-cut graph with a known unique optimum and an odd cycle, using a fixed coupling strength and SHIL setting, for many repeated runs; if the runs give different partitions or a cut below the known optimum, the convergence-to-ground-state claim is refuted. A benchmark version of the same test is to compare the machine's outputs against the known optimal values of the G-set max-cut instances, which the paper has not yet done.","tokens_in":9212,"feed_emoji":"🎛️","tokens_out":10133,"duration_ms":83836,"temperature":0.7,"pith_summary":"This paper reports the construction and testing of an analog Ising machine built from eight phase-shift oscillators wired through a configurable matrix of digital potentiometers. The authors' aim is to show that a non-quantum, off-the-shelf oscillator network can solve max-cut problems by letting the oscillators synchronize into binary phases that encode the partition of the graph. They find that when a second-harmonic injection-locking (SHIL) signal is applied at twice the oscillator frequency, the phases reliably binarize and the machine consistently produced accurate solutions for the small max-cut graphs they designed. The result matters because it suggests a cheap, parallel, analog route to combinatorial optimization that avoids both digital simulation cost and quantum hardware. The paper is explicit that the tested problems were author-designed and that coupling strengths had to be tuned per problem, with odd cycles sometimes landing in different partitions on repeated runs.","feed_headline":"Eight analog oscillators solve small max-cut graphs","feed_subtitle":"Cheap off-the-shelf parts find max-cut partitions by locking oscillator phases into 0 and π.","key_machinery":"The carrying mechanism is a network of eight phase-shift oscillators, each an operational amplifier with three RC stages, coupled through a zero-diagonal 8 by 8 matrix of 10-bit digital potentiometers whose resistances encode the graph's adjacency matrix. The computation is the synchronization dynamics: oscillators connected by an edge tend to settle $\\pi$ out of phase, unconnected ones in phase, following Kuramoto-type phase coupling. A second-harmonic injection-locking (SHIL) signal, an external signal at twice the resonance frequency, is added to every oscillator's synchronization input to force the continuous phases to binarize into $0$ or $\\pi$, making the final state readable as a graph partition. The phase detectors multiply each oscillator's output against the reference and integrate, giving a clean binary readout of the solution.","core_discovery":"On the paper's own terms, the central discovery is that an oscillator-based Ising machine can be assembled from standard analog computer modules and purpose-built phase-shift oscillators, and that this machine solves small max-cut instances by antiphase synchronization. Each node of the graph is an oscillator; a cut edge is represented by two oscillators settling into opposite phases, $0$ and $\\pi$, an uncut edge by in-phase synchrony, and the assignment of phases to the two sides of the partition is read out directly. The mechanism that makes this reliable is an external SHIL signal at twice the oscillator resonance frequency, which forces the phases to binarize; with such a signal at 2.5 V and 6.8 kHz, using sinusoidal, rectangular, or pulse waveforms, the machine 'consistently produced accurate solutions' for the problems tested. The authors also report that coupling strength must be chosen per problem, that odd closed loops can produce different partitions on repeated runs, and that isomorphic graphs give identical results under SHIL.","pith_inferences":["I infer that the reported accuracy is conditional on per-instance parameter selection, since the authors say each problem needed its own coupling strength and SHIL choice; without a fixed setting that works across a benchmark family, the machine is not yet a drop-in solver.","I infer that the odd-loop instability is the main scaling risk: any graph with an odd cycle has competing phase assignments, and if repeated runs sample among them, a single run's answer needs independent verification.","The paper has not yet run the G-set benchmarks it names; I infer that a passing result on those graphs without per-graph tuning would be the decisive test of the central claim.","Because the 10-bit weights were only lightly exercised, I infer that much coarser couplings, possibly ternary or single-bit weights, may suffice, which would simplify the interconnect for larger machines."],"forward_implications":["A max-cut graph maps directly to the hardware: the adjacency matrix sets the resistive couplings, and the converged binary phases read out as the partition.","With a SHIL signal at twice the oscillator frequency, the machine consistently produced accurate solutions for the author-designed max-cut graphs tested, with no digital search over partitions.","Different graphs require different coupling strengths, with values such as 1/10 or 1/5 of full scale working for most, so the hardware is not yet parameter-free.","Graphs with odd closed loops can synchronize into different partitions on repeated runs, so the dynamics do not always settle into a unique ground state.","Driving the SHIL at three times the oscillator frequency makes the phases ternarize, which the paper suggests could allow states \\{-1,0,1\\} and greater expressive power."],"supporting_citations":[{"why":"Gives the oscillator-Ising mapping and the SHIL-based binarization that the machine's readout relies on.","marker":"WANG et al. (2017)"},{"why":"Supplies the synchronization theory behind the claim that coupled oscillators converge toward the ground state.","marker":"PIKOVSKY et al. (2001)"},{"why":"Reviews Ising machines as hardware solvers, positioning the experiment as an instance of that approach.","marker":"MOHSENI et al. (2022)"},{"why":"Provides the Ising-model formulation and the link between energy minimization and max-cut.","marker":"BIAN et al. (2010)"},{"why":"Cited for the synchronization behavior of graphs with odd closed loops and for isomorphic ring-graph results.","marker":"BELYKH et al. (2005)"},{"why":"Cited for parameter tuning of such machines and for multistable states in coupled phase oscillators.","marker":"YUAN et al. (2017)"},{"why":"Provides the Kuramoto model and general synchronization concepts used in the paper's mathematical discussion.","marker":"STROGATZ (2018)"},{"why":"Cited on using noise to augment synchronization, the basis for the paper's open question about escaping local minima.","marker":"VAIDYA et al. (2021)"}],"fun_headline_variants":["Analog oscillator Ising machine solves max-cut graphs","Phase-locked oscillators solve small max-cut instances","Off-the-shelf analog parts build an Ising max-cut machine","Oscillator Ising machine uses SHIL to solve max-cut"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the network settles into the true ground state of the Ising Hamiltonian, the actual max-cut, whenever coupling weights and SHIL parameters are set, rather than into some local minimum; this is not proven, and the paper itself notes that odd closed loops often produce different partitions on repeated runs.","fun_headline_variants_meta":{"raw":{"variants":["Analog oscillator Ising machine solves max-cut graphs","Phase-locked oscillators solve small max-cut instances","Off-the-shelf analog parts build an Ising max-cut machine","Oscillator Ising machine uses SHIL to solve max-cut"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000464,"raw_usage":{"total_tokens":2283,"prompt_tokens":877,"completion_tokens":1406,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":493,"completion_tokens_details":{"reasoning_tokens":1338}},"tokens_in":493,"tokens_out":1406,"duration_ms":10941,"temperature":1.0,"reasoning_tokens":1338,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T05:39:15.550232+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the machine on an 8-node max-cut graph with a known unique optimum and an odd cycle, using a fixed coupling strength and SHIL setting, for many repeated runs; if the runs give different partitions or a cut below the known optimum, the convergence-to-ground-state claim is refuted. A benchmark version of the same test is to compare the machine's outputs against the known optimal values of the G-set max-cut instances, which the paper has not yet done.","supporting_citations":[],"review_version":1}