{"id":"74adbffd-f3c4-4ca3-a5c6-2afaa1f18d46","arxiv_id":"2611.17061","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Multi-particle quantum walk search on a 2D grid achieves a constant-factor improvement in runtime over the single-particle walk, but retains the O(√N log N) scaling.","lead":"This paper studies multi-particle quantum walk search on a two-dimensional grid, finding that adding particles yields a constant-factor speedup over the single-particle case but does not change the asymptotic O(√N log N) scaling. The work provides a generalized interaction framework and numerical simulations for two and three walkers.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"The two-level approximation (Eq. 39) is not analytically justified for P>1, and the claim that τ(P) ≃ τ(1)/√P for P=2 is extracted from a fit whose validity for multi-walker dynamics is unproven.","rationale":"The reader's weakest_assumption correctly identifies that the two-level approximation is not analytically justified for multiple particles. The stress-test agrees that this is the most load-bearing concern. The paper's central claim of a constant-factor speedup is extracted from a fit that assumes this approximation. Without a proof or a numerical check that the multi-particle dynamics are indeed dominated by two eigenvectors, the fitted τ is a phenomenological parameter whose relation to the search runtime is unclear. The concrete test proposed (exact diagonalization for small lattices) would directly verify the two-level assumption. The verdict remains CONDITIONAL because the paper is otherwise well-structured and the numerical evidence for P=2 is extensive, but the concern is significant enough that the claim cannot be accepted without addressing it. The agreement is 'agree' because the reader and stress-test identify the same weakest assumption.","tokens_in":24819,"tokens_out":784,"duration_ms":6473,"concrete_test":"Compute the eigenphases of U' (Eq. 13) for P=2 on small lattices (e.g., L=5,7,9) by exact diagonalization. If the two-level approximation holds, the two dominant eigenphases should have a gap Δ = π/τ that scales as 1/√(N log N) and the initial state should have overlap >0.99 with the subspace they span. If the gap does not follow this scaling or the overlap is small, the fitted τ does not correspond to a genuine two-level search and the speedup claim is unsupported.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central quantitative claim—that the multi-particle search achieves a constant-factor speedup τ(P) ≃ τ(1)/√P—rests entirely on fitting the success probability to the two-level form p_succ(T) ≃ A sin²(πT/(2τ)) (Eq. 39). For a single walker this form is justified by the analysis of Ref. [15], which shows that the dynamics are dominated by two eigenvectors of the evolution operator. For P>1 the authors explicitly state that they 'treat it as a model and test it against the simulations' (Sec. VII). The test consists of checking that the fit improves with lattice size, but no analytical argument is given that the multi-particle evolution operator U' (Eq. 13) has a dominant two-dimensional eigenspace. Without such an argument, the extracted τ is a phenomenological parameter whose relation to the actual search runtime is unclear. The claim that τ(P)/τ(1) → 1/√P is then a statement about this fitted parameter, not about the operational peak time T_peak, which the authors themselves note is 'nearly degenerate' for P=1 and 'better defined' for P=2 (Fig. 11). The ratio of fitted τ values may reflect the fitting procedure rather than a genuine speedup. The paper acknowledges that the two-level approximation is 'not proven analytically' (reader's weakest_assumption), but the stress-test must note that the entire speedup claim depends on this unproven assumption. The concrete test would be to derive the effective Hamiltonian for the multi-particle walk and check whether the spectral gap between the two dominant eigenphases indeed scales as 1/√(N log N) with a √P prefactor, or to compute the overlap of the initial state with the putative two-level subspace.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"The paper studies multi-particle quantum walk search on a 2D lattice, focusing on the any-walker criterion. It constructs a joint marked coin operator that couples the walkers, defines phase and HPP interaction models, and simulates the dynamics for P=1,2,3. The central claim is that the multi-particle search achieves a constant-factor speedup τ(P) ≃ τ(1)/√P while preserving the O(√N log N) scaling. The paper also provides a distributed halo-exchange algorithm for larger simulations and analyzes entanglement generation.","tokens_in":25247,"tokens_out":2456,"duration_ms":18674,"significance":"If the constant-factor speedup holds, it is a practically relevant improvement for quantum walk search on 2D lattices, which is the critical dimension where the single-particle walk does not reach Grover's bound. The paper also provides a formal generalization of interaction models and a distributed simulation method. However, the central quantitative claim rests on a fit to the two-level approximation, which is not analytically justified for P>1, and the evidence for P=3 is limited to very small lattices.","major_comments":[{"comment":"The entire speedup claim τ(P) ≃ τ(1)/√P is extracted from fitting the success probability to the two-level form p_succ(T) ≃ A sin²(πT/(2τ)). For a single walker this form is justified by the analysis of Ref. [15], but for P>1 the authors state that they 'treat it as a model and test it against the simulations' (Sec. VII). The test consists of checking that the fit improves with lattice size, but no analytical argument is given that the multi-particle evolution operator U' (Eq. 13) has a dominant two-dimensional eigenspace. Without such an argument, the extracted τ is a phenomenological parameter whose relation to the actual search runtime is unclear. The claim that τ(P)/τ(1) → 1/√P is then a statement about this fitted parameter, not about the operational peak time T_peak, which the authors themselves note is 'nearly degenerate' for P=1 and 'better defined' for P=2 (Fig. 11). The ratio of fitted τ values may reflect the fitting procedure rather than a genuine speedup. The paper acknowledges that the two-level approximation is 'not proven analytically', but the entire speedup claim depends on this unproven assumption.","section":"§VII, Eq. (39)","authors_acknowledge":"It is important to note, however, that since the walk operator is applied an integer number of times,τserves only to characterize the scaling, while Tpeak andp peak remain the operational figures of merit at each simulated size.","authors_acknowledge_p":0.7},{"comment":"The evidence for P=3 is extremely limited: the simulated range is only L=5 to L=25 (N=25 to 625). The ratio τ(1)/τ(3) in Fig. 13 has not converged and is still far from √3 ≈ 1.732. The authors themselves state that 'the simulated lattices are too small to determine the limit' (Sec. VII). The claim that the constant-factor speedup extends to P=3 is therefore unsupported by the data presented. The paper should either provide larger-scale simulations for P=3 (perhaps using the distributed algorithm with more resources) or clearly state that the P=3 result is preliminary and not conclusive.","section":"§VII, Fig. 13","authors_acknowledge":"ForP= 3, the simulated lattices are too small to determine the limit.","authors_acknowledge_p":0.94},{"comment":"The conclusion that local interactions do not change the runtime (T_peak is independent of α and λ) is based on simulations up to L=16 for P=2 and L=10 for P=3. These are very small lattices where finite-size effects are strong. The authors then state that 'we therefore did not need to study how the runtime scales with the database size in the presence of these interactions' (Sec. VIII). This is a strong claim to base on such limited data. It is possible that interactions affect the scaling prefactor or even the asymptotic scaling at larger sizes, and the small-lattice simulations are not sufficient to rule this out.","section":"§VI, Fig. 9","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The peak success probability for P=2 is shown to match or slightly exceed the single-walker value for N≳900. However, the data for P=2 only go up to L=230 (N≈53,000). The trend in Fig. 10 (right panel) shows that p_peak for P=2 is still fluctuating and has not clearly converged relative to P=1. The claim that 'from N≈900 onwards it matches or slightly exceeds the single-walker value' is based on a few data points and should be stated more cautiously.","section":"§V, Fig. 10","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The definition of the joint perturbed coin C' in Eq. (12) is given in terms of tensor factors reordered as H_c^⊗P ⊗ H_p^⊗P. The notation is somewhat confusing; it would be clearer to write C' = G^⊗P ⊗ (I_p^⊗P - P_m) - I_c^⊗P ⊗ P_m, where P_m is the projector onto configurations with at least one walker at m. The current expression with P_m defined in Eq. (11) is correct but hard to parse.","section":"§III B, Eq. (12)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The phase interaction model assigns three phases λ_∥, λ_⇄, λ_⊥ to parallel, antiparallel, and perpendicular joint direction states. The paper then restricts to the uniform phase λ_∥ = λ_⇄ = λ_⊥ = λ in §VI B. It would be helpful to state explicitly that the uniform phase is a special case and that the general three-parameter model is not explored numerically.","section":"§III D, Eq. (18)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The halo-exchange algorithm is described clearly, but the pseudocode in Algorithm 1 applies the Grover coin to each walker sequentially (lines 2-4) and then applies the marked coin (lines 5-7). However, the marked coin C' in Eq. (12) replaces the Grover coin with -I_c on marked configurations, so the order should be: apply G to all walkers, then overwrite marked slices with -I_c. The pseudocode does this correctly, but the text could be more explicit about the fact that the marked coin is not a separate operator but a modification of the Grover coin.","section":"§IV, Algorithm 1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The definition of T_peak as 'the step at which the success probability p_succ(T) is maximal within its first period' is clear, but the paper later notes that for P=1 the maximum is nearly degenerate (Fig. 11). It would be useful to state the criterion used to select the maximum when there are multiple steps with nearly equal values (e.g., the first such step, or the one with the highest value).","section":"§V A","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The plot of τ²/N vs ln N is a good way to visualize the scaling. However, the linear fits are shown only for L≥50 for P=1 and P=2. The fit range and the fit parameters (slope, intercept, R²) should be reported in the text or a table to allow quantitative comparison.","section":"§VII, Fig. 12","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The introduction states that 'the first references reported a quantum walk search algorithm back in the early 2000s' and cites Refs. [14, 15]. Ref. [14] is Shenvi, Kempe, Whaley (2003) and Ref. [15] is Ambainis, Kempe, Rivosh (2004). The AKR paper is indeed a landmark, but the Shenvi-Kempe-Whaley paper is also important and should be cited correctly. The current citation [14] is to the Shenvi paper, but the text seems to attribute the AKR algorithm to both.","section":"§I, Introduction","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The discussion of the classical hitting time on a 2D grid (Θ(N log N)) is helpful for intuition. However, the statement that 'P independent classical walkers reduce the hitting time by roughly a factor of P' is not precise: the hitting time for the fastest of P independent walkers scales as Θ(N log N / P) only if the walkers are independent and the search is for the first walker to hit the target. This is a standard result, but a brief justification or reference would be useful.","section":"§II","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The interaction operator V_ab is defined as I + (O_ab - I_c^⊗2) ⊗ Π_ab. This is correct, but it might be helpful to note that this is a controlled unitary: if the positions coincide, apply O_ab; otherwise, do nothing.","section":"§III C, Eq. (15)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The entropy S(ρ_1) is plotted in bits (log base 2). The bound log₂(4N) ≈ 7.6 bits is mentioned. It would be useful to also plot the bound as a horizontal line in Fig. 7 for reference, as is done in Fig. 8.","section":"§V C, Fig. 7","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The uniform phase interaction results show symmetry about λ=π, which the authors attribute to complex conjugation. This is correct, but it would be helpful to note that the symmetry holds only because the initial state and all operators are real. If a complex phase were introduced elsewhere, the symmetry would break.","section":"§VI B, Fig. 9(b,d)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The conclusion states that 'the geometry of the grid preserves the O(√N log N) scaling of the runtime'. This is consistent with the data for P=2, but for P=3 the data are insufficient to confirm the scaling. The statement should be qualified.","section":"§VIII, Conclusions","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Ref. [19] (Mittal and Sowiński) is cited as 'Physical Review A 113, 10.1103/tg6f-v65l (2026)'. The DOI appears to be a placeholder or an internal identifier. Please verify that the reference is correct and accessible.","section":"References","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision"},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"The paper does two things well. First, it constructs a genuine multi-particle quantum walk search algorithm with a joint marked coin that respects the any-walker criterion, which is not the same as running P independent single-particle searches. Second, it provides a clean numerical demonstration that for P=2 the fitted runtime τ(2) is smaller than τ(1) by a factor approaching √2 as N grows, and that the asymptotic scaling remains O(√N log N). The distributed halo-exchange simulation code is a practical contribution. The paper is clearly written and the limitations are stated openly.\n\nThe soft spot is the two-level approximation that underlies the runtime extraction. For a single walker the form p_succ(T) ≃ A sin²(πT/(2τ)) is justified by the spectral analysis of Ref. [15]. For P>1 the authors treat it as a model and test it numerically, but the fit quality improving with lattice size is not a proof that the multi-particle evolution operator has a dominant two-dimensional eigenspace. The extracted τ is a phenomenological parameter; the claim that τ(P) ≃ τ(1)/√P is a statement about this fitted parameter, not about the operational peak time T_peak, which for P=1 is nearly degenerate and for P=2 is better defined. The paper acknowledges this limitation, but the stress-test concern is real: the entire speedup claim depends on an unproven assumption. The P=3 data is too limited (L up to 25) to confirm the √P trend, and the authors say so.\n\nThat said, the paper is honest about what it does and does not prove. The constant-factor speedup for P=2 is a genuine numerical result, even if its interpretation depends on the fit model. The paper does not claim to have resolved the open question of O(√N) scaling on the 2D grid.\n\nWho this is for: researchers working on quantum walk search algorithms, especially those interested in multi-particle extensions and numerical simulation techniques. A serious referee should engage with the paper, but should push the authors to either provide an analytical argument for the two-level approximation for P>1 or to present the operational peak time T_peak as the primary figure of merit alongside the fitted τ. The paper deserves peer review.","headline":"Solid numerical demonstration of a constant-factor speedup for P=2 quantum walk search on a 2D grid, but the central claim rests on an unproven two-level approximation and the P=3 data is too limited to confirm the √P trend.","tokens_in":25822,"tokens_out":582,"duration_ms":6913},"pith_extraction":{"msc":["81P68","81V70","68Q12"],"pacs":["03.67.Ac","03.67.Lx","05.40.Fb"],"model":"deepseek/deepseek-v4-flash","headline":"Multi-particle quantum walk search on a two-dimensional lattice achieves constant-factor speedup","keywords":["quantum walk","multi-particle quantum walk","spatial search","two-dimensional grid","constant-factor speedup","local interactions","halo exchange simulation"],"falsifier":"A direct analytic calculation of the eigenvalue gap of U' for P=2 on an infinite 2D grid would either confirm the √2 factor or show that the constant factor deviates from √P. Alternatively, a simulation of P=2 on a significantly larger lattice (e.g., L=500) with higher precision could test whether the ratio τ(1)/τ(2) continues to approach √2 or saturates at a different value.","tokens_in":26279,"feed_emoji":"🔍","tokens_out":1611,"duration_ms":10119,"temperature":0.7,"pith_summary":"The paper studies whether adding extra particles to a quantum walk search on a 2D grid can improve the search runtime. It defines a multi-particle quantum walk search algorithm that uses a joint marked coin operator, ensuring that the presence of any walker at the target vertex triggers the oracle. The authors simulate the algorithm numerically for two and three walkers on lattices up to size 230×230 (two walkers) and 25×25 (three walkers). They find that the asymptotic scaling remains O(√N log N), the same as the single-particle case, but the runtime is reduced by a constant factor that approaches √P for large N. Local interactions (phase and HPP collisions) generate additional entanglement but do not change the runtime or improve the success probability. The paper provides a new generalized framework for local interactions among P walkers and a distributed simulation algorithm using halo exchange.","feed_headline":"Extra particles speed up 2D quantum walk search by √P","feed_subtitle":"Multi-particle quantum walk on a grid achieves a constant-factor runtime improvement without changing the asymptotic scaling.","key_machinery":"The central object is the joint multi-particle marked coin operator C' (Eq. 12), which applies a π phase shift to every configuration where at least one walker is at the marked vertex. This replaces the tensor product of independent marked coins and ensures the any-walker success criterion. The evolution operator is U' = S_ff^⊗P C'. The analysis also relies on the sinusoidal fit p_succ(T) ≈ A sin²(πT/(2τ)) (Eq. 39), which holds when the search dynamics are dominated by two eigenvectors of U', and the fitted τ extracts the runtime.","core_discovery":"The paper establishes that multi-particle quantum walk search on a 2D grid, using the any-walker criterion and a joint marked coin operator, achieves a constant-factor improvement in runtime over the single-particle case. The ratio τ(1)/τ(P) approaches √P for large N, as confirmed numerically for P=2 and suggested for P=3. The asymptotic scaling remains O(√N log N), so the logarithmic overhead of the 2D grid is not removed. The paper also shows that local interactions (phase and HPP) generate additional entanglement but do not change the runtime or improve the success probability, so the non-interacting search is the relevant one for scaling analysis.","pith_inferences":["The constant-factor speedup √P is reminiscent of the classical speedup from P independent random walkers, but the quantum walkers are entangled through the joint coin, so the mechanism is genuinely quantum. This suggests that multi-particle quantum walks may achieve similar constant-factor improvements in other search and hitting-time problems on recurrent graphs.","The numerical evidence for P=3 is limited to small lattices (L ≤ 25), so the √3 extrapolation is tentative. A more efficient simulation method or an analytic argument would be needed to confirm the trend for larger P.","The paper's framework for local interactions among P walkers (Eq. 16) could be extended to study other collective phenomena in quantum walks, such as topological phases or many-body localization, where interactions play a central role.","The finding that interactions do not affect the runtime but do affect the success probability suggests that the search is robust to certain types of noise or decoherence that mimic local phase shifts, which could be relevant for experimental implementations."],"forward_implications":["Multi-particle quantum walk search can be directly implemented on quantum simulator platforms where nearest-neighbor locality is a constraint, offering a constant-factor speedup without changing the underlying hardware.","The √P reduction in runtime means that adding more walkers yields diminishing returns (e.g., four walkers halve the steps compared to one), but the practical benefit depends on the physical cost of adding particles.","The joint marked coin operator provides a template for multi-particle quantum search algorithms beyond walks, e.g., for multi-query Grover search or distributed quantum sensing.","The halo-exchange simulation algorithm enables numerical study of multi-particle quantum walks on larger lattices than previously possible, which may be useful for benchmarking other multi-particle quantum algorithms.","The observation that local interactions do not improve the search suggests that engineering interactions is not a promising route to overcome the 2D logarithmic overhead; other modifications (e.g., Tulsi's ancilla) may be needed."],"supporting_citations":[{"why":"Establishes the single-particle quantum walk search algorithm on a 2D grid, including the O(√N log N) scaling and the sinusoidal fit (Eq. 39) that the paper generalizes to multiple particles.","marker":"[15]"},{"why":"Provides Tulsi's algorithm for faster 2D spatial search using an ancilla qubit, which the paper suggests as a future direction for multi-particle search.","marker":"[16]"},{"why":"Supplies the two-walker phase interaction framework that the paper generalizes to P walkers.","marker":"[19]"},{"why":"Introduces the HPP (Hardy–Pomeau–de Pazzis) collision model for multi-particle quantum walks, which the paper adapts as a tunable interaction.","marker":"[18]"},{"why":"Grover's algorithm provides the lower bound for unstructured search, which the paper uses as a reference for the optimal scaling.","marker":"[20]"},{"why":"Szegedy's quantum walk framework connects hitting times to search, motivating the multi-particle analysis via classical hitting-time reduction.","marker":"[21]"},{"why":"Establishes the critical dimensionality of the grid for quantum walk search, explaining why 2D is the focus.","marker":"[23]"}],"fun_headline_variants":["Multi-particle quantum walk speeds grid search by √P","Two walkers beat one: constant-factor speedup in 2D quantum search","Adding particles to a quantum walk search gives constant factor gain","Quantum walk search on 2D grid gets √P boost from extra walkers","Multi-particle search on 2D lattice: constant factor improvement"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The analysis assumes that the search dynamics are dominated by two eigenvectors of the evolution operator U', leading to the sinusoidal fit of Eq. (39). This is proven for the single-particle case by Ref. [15], but for multiple particles it is treated as a model and tested only numerically; the fit quality improves with lattice size, but the assumption is not proven analytically.","fun_headline_variants_meta":{"raw":{"variants":["Multi-particle quantum walk speeds grid search by √P","Two walkers beat one: constant-factor speedup in 2D quantum search","Adding particles to a quantum walk search gives constant factor gain","Quantum walk search on 2D grid gets √P boost from extra walkers","Multi-particle search on 2D lattice: constant factor improvement"]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":7.8e-05,"raw_usage":{"cost":7.79142e-05,"is_byok":false,"cost_details":{"upstream_inference_cost":7.79142e-05,"upstream_inference_prompt_cost":6.17162e-05,"upstream_inference_completions_cost":1.6198e-05},"total_tokens":972,"prompt_tokens":883,"completion_tokens":89,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":256,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":0}},"tokens_in":883,"tokens_out":89,"cost_source":"openrouter_usage","duration_ms":2166,"image_pages":[],"temperature":1.0,"billed_attempts":1,"reasoning_tokens":0,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-13T11:47:09.160428+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"pipeline_version":"pith-pipeline@v1.4.1-daily-deepseek-or-medium","input_kind":null,"input_chars":null,"input_total_chars":null,"passes_missing":null,"published_standing":null,"published_confidence":null,"objection_audit":null,"run_record":null,"basis":{"kind":"not_recorded","source":null,"chars_given":null,"chars_total":null},"receipt":null,"falsifier":"A direct analytic calculation of the eigenvalue gap of U' for P=2 on an infinite 2D grid would either confirm the √2 factor or show that the constant factor deviates from √P. Alternatively, a simulation of P=2 on a significantly larger lattice (e.g., L=500) with higher precision could test whether the ratio τ(1)/τ(2) continues to approach √2 or saturates at a different value.","supporting_citations":[{"cited_title":"Coins Make Quantum Walks Faster","cited_arxiv_id":"quant-ph/0402107","evidence_quote":"Establishes the single-particle quantum walk search algorithm on a 2D grid, including the O(√N log N) scaling and the sinusoidal fit (Eq. 39) that the paper generalizes to multiple particles."},{"cited_title":"Faster quantum-walk algorithm for the two- dimensional spatial search, Phys. Rev. A78, 012310","cited_arxiv_id":null,"evidence_quote":"Provides Tulsi's algorithm for faster 2D spatial search using an ancilla qubit, which the paper suggests as a future direction for multi-particle search."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the two-walker phase interaction framework that the paper generalizes to P walkers."},{"cited_title":"Multiparticle quantum walk with a gaslike interaction, Phys. Rev. A 100, 042320","cited_arxiv_id":null,"evidence_quote":"Introduces the HPP (Hardy–Pomeau–de Pazzis) collision model for multi-particle quantum walks, which the paper adapts as a tunable interaction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Grover's algorithm provides the lower bound for unstructured search, which the paper uses as a reference for the optimal scaling."},{"cited_title":"Szegedy, Quantum speed-up of markov chain based algorithms, in45th Annual IEEE Symposium on Foun- dations of Computer Science (FOCS)(2004) pp","cited_arxiv_id":null,"evidence_quote":"Szegedy's quantum walk framework connects hitting times to search, motivating the multi-particle analysis via classical hitting-time reduction."},{"cited_title":"Spatial search and the 17 dirac equation, Phys. Rev. A70, 042312","cited_arxiv_id":null,"evidence_quote":"Establishes the critical dimensionality of the grid for quantum walk search, explaining why 2D is the focus."}],"lane_accuracy":{"status":"not yet measured","lane":{"model":"deepseek/deepseek-v4-flash","input_kind":null},"url":"https://pith.science/accuracy"},"review_version":1}