Papers

Large language models (LLMs) have recently been applied in systems research as a tool to reduce human-intensive engineering effort through cost-efficient automation. Decades of research have produced a rich landscape of concurrency control (CC) protocols, each encoding distinct trade-offs in correctness, throughput, and abort behavior. However, most applications in practice default to 2PL or OCC, because selecting and adapting a protocol to a specific application requires expert knowledge that is rarely available to application designers. This is a wasted opportunity, as an application-specific CC protocol can yield significant performance advantages over a generic baseline, but designing one requires deep expertise in CC protocol design. In this paper, we propose HakiCC, an LLM-driven multi-agent pipeline that automatically designs, verifies, and optimizes concurrency control protocols tailored to a given target application. HakiCC provides a two-stage pipeline. In Stage 1, a multi-agent system takes a workload description as input and generates an application-specific CC protocol implementation, which is iteratively repaired and verified for conflict-serializability. In Stage 2, the verified protocol is further optimized for that application through an LLM-driven evolutionary loop targeting correctness and throughput. We evaluate HakiCC on TPC-C and AuctionMark as target workloads, producing and reporting ten application-specific CC protocols. All ten are conflict-serializable after Stage 1; Stage 2 improves throughput for every protocol, with average gains of +50.6% for TPC-C protocols and +92.2% for AuctionMark protocols.

Multi-token prediction (MTP) improves the throughput of autoregressive generation by enabling the language model to draft multiple next tokens per forward pass, while a verification step over draft tokens ensures that token distribution of the backbone is preserved. Every open MTP-family release (MiMo-7B, DeepSeek-V3, Qwen3) trains its heads jointly with the backbone over the full pretraining run of tens of trillions of tokens, thus setting the drafter quality at pretraining time. We ask whether a lightweight post-training pass on target-generated chain-of-thought is enough to reach the same expected throughput speedup on a frozen reasoning model, and study how a serving-time system built on such a checkpoint can be optimized. We present three findings. 1) On a frozen Qwen3-8B with $K{=}3$ chained MTP heads, we show that a post-training recipe with plain cross-entropy on $\approx\!2.5$B tokens reaches or exceeds the expected speedup of jointly trained MiMo-7B on math, coding and knowledge benchmarks. Our post-training recipe utilizes $10^3$-$10^4\times$ less MTP-training tokens as compared with joint pre-training of MiMO-7B MTP baseline. 2) We propose a chain-aware relaxation of draft token verification rule that allows a bounded drift from backbone language model token distribution. We show that this relaxation lifts expected speedups by $+12$ to $+16\%$ per benchmark while preserving task accuracy. 3) We propose an adaptive controller that dynamically chooses the number of MTP heads to be engaged at inference time and demonstrate recovery of upto $11$–$14\%$ loss in speedup using fixed maximum MTP draft length.

Codon optimization is a challenging combinatorial optimization problem with important applications in synthetic biology, protein expression, and biotechnology. While quantum annealing has previously been explored for this problem, the application of gate-based quantum algorithms remains largely unexplored. In this work, we present a gate-based quantum computing framework for codon optimization by formulating the optimization objective as an Ising Hamiltonian and expressing it in terms of Pauli operators suitable for gate-based quantum processors. The proposed framework is investigated using the Variational Quantum Eigensolver (VQE), Sampling Variational Quantum Eigensolver (SVQE), and the Quantum Approximate Optimization Algorithm (QAOA), with benchmarking against exact diagonalization, classical optimization, and quantum annealing. Using a three-amino-acid benchmark, the proposed implementations successfully reproduce the exact ground-state solutions while identifying optimized ansatz selections, optimization strategies, initialization schemes, and hardware execution settings that provide reproducible performance on current noisy quantum hardware. Hardware demonstrations are performed on IBM Quantum processors to evaluate the practical implementation of the proposed workflow. As a work in progress, this study establishes a reproducible computational framework for gate-based quantum codon optimization and provides the foundation for future investigations of larger protein sequences as quantum hardware and algorithms continue to advance.

We prove that a local minimal hypersurface $M^3$ in the unit sphere $S^4$ with constant scalar curvature and three distinct principal curvatures at every point must be locally congruent to the minimal Cartan isoparametric hypersurface in $S^4$, which provides a positive answer to the Bryant conjecture \cite{ChangLocal}. The key ingredient is to prove that, for each fixed \(S_0>3\), the set $\mathscr{G}_{S_0}$ of fourth-order jets of normalized local graph solutions satisfying \(|A|^2\equiv S_0\) is compact, where \(A\) denotes the shape operator.

Coding agents increasingly automate Lean proof development, but successful compilation alone does not establish that a candidate proves the intended statement under acceptable assumptions. We present FORALL-LEAN-AGENT, a frontend-agnostic framework for auditable reasoning in formal mathematics and software verification. The framework combines isolated workspaces, Lean tools, and fresh review with statement comparison, axiom audits, and independent proof checking where supported. Verification evidence and reviewer decisions are bound to the same candidate artifact, making acceptance traceable. We evaluate the framework on VeriSoftBench, PutnamBench, and both problems in the Lean Eval softwareverification track. On the 100-task VeriSoftBench subset, integration with FORALLLEAN-AGENT raises benchmark-rule success from 93 to 100 for GPT-5.6 Sol at low effort while reducing cost from $69 to $62. The PutnamBench evaluation accepts all 672 problems at an average of $4.72 each. These results show that agent harness design can improve correctness and efficiency while providing evidence beyond aggregate solve counts.

Objective: This work proposes a dynamic state-space statistical model of heartbeat dynamics whose optimal state estimate can be solved efficiently with convex optimization and used to assess autonomic and cardiovascular health. Methods: This state-space model combines a Gauss-Markov process prior with gamma emissions using a log-link function. Parameters are estimated using a static version of the model, and alternating direction method of multipliers is used to efficiently find the optimal maximum a priori latent state estimate. Model selection is performed by finding the tradeoff between model size and model fit using the partial autocorrelation function. Results: We demonstrate in synthetically generated heartbeats that this model is able to closely recover the underlying latent weights. Then, we show using a tilt table dataset that this statistical model is able to better capture sharp transients in heart rate patterns than traditional local averaging methods. In addition, combining heart beat timings with information from other organs such as the respiratory phase, we show that our statistical modeling framework enables the extraction of information-theoretic measures to quantify inter-organ coupling such as respiratory sinus arrythmia. Conclusion: This statistical framework for modeling heartbeat dynamics provides an efficient framework that can provide physiologically meaningful insights. Significance: This work broadens potential tools that can be used to assess heartbeat dynamics.

Robust detection of AI-generated text under deployment conditions is challenging: distribution shifts across domains and generators, adversarial perturbations of the input surface, and the absence of target-domain labels for threshold calibration all degrade detectors that perform well in-domain. We present DeBERTa-ConPara, a deployment-oriented detector combining attack-aware Unicode preprocessing with a contextual transformer encoder trained over HC3 Plus, M4, MAGE and RAID. Our central finding is that preprocessing acts in opposite directions depending on where it is applied: normalising the training corpus deduplicates it, collapsing 35.4% of RAID rows into copies of their clean siblings and deleting the adversarial supervision, whereas normalising at inference is an effective defence. A factorial varying the two placements independently identifies raw training with normalised inference as the best configuration, reaching 99.61% AUROC, 99.01% TPR@5% FPR and 96.57% TPR@1% FPR on the official RAID hidden test, alongside 93.14% average balanced accuracy across HC3 Plus and MAGE under a fixed threshold. The gain is confined to two of twelve attack classes: homoglyph and zero-width-space insertion rise from 11.05% and 1.12% to 96.98%. The same signature reproduces in a zero-shot detector of different architecture, showing the effect belongs to the attacks rather than to our model. We additionally report two negative results: semantic-invariance augmentation through paraphrasing and supervised contrastive learning (ConPara) does not improve the best configuration, and the handcrafted feature-fusion branch is inert in distribution and harmful outside it.

Iterative methods are extremely popular for solving linear ill-posed inverse problems. While deterministic convergence, or more precisely, semi-convergence of such methods is widely studied, the problem of quantifying uncertainty in the resulting reconstructions due to random noise in the observed data has received far less attention. In this work, we study uncertainty quantification for the Landweber iteration in linear inverse problems using the Radon transform as a motivating example. We interpret the Landweber reconstruction statistically by analyzing how uncertainty in the measured data is propagated through the reconstruction map. This perspective is closely related to generalized fiducial inference, where uncertainty about the parameter is induced by inverting the relation between the observed data and the unknown (fixed) quantity. We combine the resulting stochastic uncertainty with a theoretical bound on the regularization bias to construct confidence intervals for the reconstructed solution. The evaluation of the resulting uncertainty estimates require repeated operations with the forward operator that can become expensive in large-scale problems. To reduce this cost, we use randomized singular value decomposition to obtain a low-rank approximation of the normal operator associated with the Radon transform and incorporate this approximation in quantifying uncertainty. {{In particular, this requires that the additional randomization error introduced due to the use of such randomized techniques be included in our approach for uncertainty quantification.}} Numerical results show that the proposed method provides accurate reconstructions and reliable uncertainty estimates at substantially reduced computational cost.

Sign language machine translation has progressed substantially over the past decade, evolving from isolated sign recognition to end-to-end translation systems. Advances in pose estimation, transformer architectures, and large-scale dataset collection have driven progress, yet challenges remain. Datasets are limited compared to spoken-language resources; evaluation metrics inadequately capture the linguistic quality of output; and models must capture the simultaneous, multi-layered, and three-dimensional structure of sign languages. This manuscript provides a comprehensive review that seeks to balance technical challenges with stakeholder considerations. We examine the linguistic properties that make sign languages computationally unique, trace the evolution of recognition, translation, and production systems, and analyze ongoing technical challenges. Crucially, we address ethical considerations around data governance, community involvement, and appropriate use. Drawing on interdisciplinary perspectives spanning computer vision, sign language linguistics, and deaf studies, our analysis emphasizes that continued progress requires sustained collaboration across these fields and with deaf communities.

In the fair dueling game, two players take turns shooting at each other, with the turn order defined by the following fairness criterion: the next shot is assigned to the player who is less likely to have won the duel so far. Cooper and Dutle showed that when both players have the same probability $p$ of a successful shot, the shooting sequence converges to the Thue–Morse sequence as $p \rightarrow 0$. We extend this model by forcing the sequence to begin with an arbitrary finite prefix before reverting to this fairness criterion. We prove that this initial disruption fundamentally alters the sequence's eventual behavior: in the $p \rightarrow 0$ limit, unless the initial prefix is itself a prefix of the Thue–Morse sequence or its complement, the resulting sequence is eventually periodic. Moreover, the least period is a power of 2 that is at most $4^{n+1}$, where $n$ is the length of the prefix, and the repeating block is itself a prefix of the Thue–Morse sequence.

We study the spectral properties of the harmonic modification of the quantum mechanical Kronig-Penney model motivated by the theory of the dressing chain. In particular, we show that in the special cases the corresponding ground states can be given explicitly and all spectral gaps have the same length with the ends expressed as zeros of the corresponding Weber-Hermite functions.

General-purpose embodied robots should support both navigation toward language-specified destinations and dynamic person tracking under arbitrary initial target azimuths. However, existing methods typically rely on forward-facing observations and address these tasks with separate policies, limiting omnidirectional perception and unified closed-loop control. We present UniTrackPLA, a unified panorama-language-action model for instruction-guided navigation and dynamic person tracking. Its Panoramic-Aware Encoding (PAE) preserves the temporal and azimuthal structure of perspective views projected from each panorama, enabling perspective-pretrained visual encoders to process omnidirectional observations. A shared vision-language backbone grounds instructions in the panoramic context and predicts continuous robot-centric waypoint chunks for both tasks. World-Action Consistency (WAC) further predicts action-conditioned future visual states and verifies waypoint prefixes online, allowing reliable actions to be reused while triggering replanning upon inconsistency. We also introduce OmniTrackNav-Bench, comprising 5,000 simulated tracking trajectories, 10,000 simulated VLN routes, and 96 verified real-world routes, providing 919,978 waypoint-supervision instances. UniTrackPLA improves overall tracking SR from 23.50% to 35.00% and Omni-VLN SR/SPL from 13.00%/12.77% to 19.75%/19.29%. Incorporating 76 real-world routes further improves held-out EP@0.2m from 42.92% to 92.08%. Closed-loop experiments on a Go2-W robot demonstrate unified panoramic tracking and navigation across indoor and outdoor environments. The project page is at https://tw5775.github.io/UniTrackPLA.

Chemical intercalation provides a powerful route for tuning the electronic and magnetic properties of transition metal dichalcogenides (TMDs), enabling the emergence of magnetic phases and phenomena that are otherwise inaccessible in their pristine forms. Here, we investigate the structural, electronic, and magnetic properties of magnetically intercalated Cr$_{1+x}$Te$_2$ through a combination of scanning tunneling microscopy, angle-resolved photoemission spectroscopy, magnetization, electrical transport, and Lorentz transmission electron microscopy measurements. We uncover a rich magnetic phase diagram comprising paramagnetic, in-plane spin-fluctuation, out-of-plane ferromagnetic, and field-induced Néel-type skyrmion phases. In particular, Lorentz transmission electron microscopy directly reveals the formation of Néel-type skyrmion textures slightly below the Curie temperature, with their configurations evolving strongly with temperature and magnetic field. The skyrmions exhibit pronounced thermal hysteresis and coexist with stripe domains, indicative of a complex magnetic energy landscape arising from the competition among disorder-induced local Dzyaloshinskii-Moriya interactions, perpendicular magnetic anisotropy, and dipolar interactions. Our results demonstrate that chemical intercalation can profoundly reshape the magnetic interactions and phase behavior of TMDs, establishing Cr$_{1+x}$Te$_2$ as a versatile platform for exploring tunable magnetism, skyrmion physics, and unconventional thermodynamic phenomena in low-dimensional quantum materials.

Machine-learned interatomic potentials are commonly fitted by weighted-sum scalarization, which combines energy and force errors in a single loss. A nominal weight, however, identifies a potential only relative to the complete fitting protocol. We therefore treat energy–force balancing as a protocol-dependent problem of physical model selection. For fixed-basis atomic cluster expansion models of molten LiCl, liquid H$_2$O, and Si, the resulting scalarization paths contain dominated states at extreme weights. Their nondominated subsets depend on the solver, and the out-of-distribution Si path is nonmonotone. We replace direct weight selection by minimizing the regularized energy objective subject to an upper bound on the normalized force loss. Projected dual ascent adjusts the Lagrange multiplier from the force-constraint residual. A frozen-multiplier limited-memory quasi-Newton refinement then returns the final model. Across the three systems, the constrained procedure reaches the solver-matched low-error region and gives measured speedups of $6.7$–$8.9$ over completed scalarization scans under the stated timing convention. Physical-property calculations show that first-shell geometry is comparatively insensitive to the selected balance, whereas transport and solid-state observables vary more strongly. These results identify the fitted model–protocol pair as the relevant object of energy–force model selection and support force-constrained fitting as an explicit selection rule.

Brandt [Discrete Math. 183 (1998) 17–25] conjectured that the least signless Laplacian eigenvalue of every regular triangle-free graph of order $n$ is at most $4n/25$. Using flag algebras, Balogh, Clemen, Lidický, Norin and Volec [SIAM J. Discrete Math. 37 (2023) 1173–1179] established the stronger bound $15n/94$ for all triangle-free graphs. We investigate the effect of additionally excluding pentagons and prove that every $\{C_3,C_5\}$-free graph $G$ of order $n$ satisfies $\qmin(G)<0.0569n$, without any regularity assumption. The proof combines seven-vertex flag inequalities with local Rayleigh constraints that retain the least eigenvalue throughout the counting argument. An exact integer certificate establishes the required inequality.

The finite group isomorphism problem asks whether two finite groups of order $N$ are isomorphic. The first algorithm, attributed to Tarjan (see Miller, STOC '78), runs in time $N^{\log N + O(1)}$. Despite intensive study, the current best known algorithm has a running time of $N^{(1 / 4 + o(1))\log N}$ (Rosenbaum, '13). $p$-groups of class $2$ have been recognized as the major bottleneck for faster group isomorphism. Recent progress has led to $N^{o(\log N)}$-time algorithms for $p$-groups of class $2$ where $p$ is odd (Sun, STOC '23; Ivanyos–Mendoza–Qiao–Sun–Zhang, FOCS '24; Grochow–Qiao–Stange–Sun, STOC '25). However, the case of $p=2$, which represents the majority of $p$-groups of class 2 assuming a well-known conjecture in group enumeration, remained elusive, with essentially no progress until now. In this paper, we present an algorithm for testing the isomorphism of two 2-groups of Frattini class 2 of order $N$ in time $N^{O((\log N)^{1/2})}$. To our knowledge, this is the first $N^{o(\log N)}$-time isomorphism algorithm for a class of $2$-groups that constitutes logarithmically almost all $2$-groups, in the sense that $\lim_{N \to \infty} \frac{\log(\text{\# 2-groups of Frattini class 2 and order } \leq N)}{\log(\text{\# 2-groups of order} \leq N)} = 1$. As our main tool, we present the first non-trivial algorithms for the quadratic form space/tuple isometry problems over $\mathbb{F}_2$. These algorithms rely on combinations of combinatorial and algebraic ideas, including finite matrix group algorithms developed by Luks (FOCS '92). As far as we know, this is the first time that matrix group algorithms are used to make progress on the worst-case complexity of $p$-group isomorphism.

In many industrial applications, 1) tabular data is scarce and imbalanced and thus requires synthetic expansion; 2) input distributions drift between training and deployment (covariate shift); 3) validation sets often diverge from unseen test environments; or 4) standard generative models simply mimic outdated source distributions. This learning setting limits the stability of standard augmentation and adaptation pipelines. We generalize the task under such setting as the Augmented and Weighted Learning under Covariate Shift problem (AWL-CS). AWL-CS imposes two critical challenges on existing methods: 1) misleading generative guidance where models optimize for source similarity rather than downstream task relevance, and 2) structural instability of distributional density where reweighting mechanisms overfit to noisy validation signals. To tackle these challenges, we propose IGDPR (Invariant-Guided Diffusion with Prototype Reweighting), a unified framework that synergizes stable synthesis and structural adaptation: i) To achieve task-relevant generation, we steer the diffusion sampling process using invariant potentials to ensure synthetic samples align with stable decision boundaries rather than outdated correlations. ii) To ensure stable adaptation, we develop a prototype-based reweighting strategy that assesses sample reliability through structural clusters instead of isolated points, effectively filtering validation noise. Extensive experiments on real data demonstrate our method improves data quality by augmenting the most beneficial data for robust learning.

Long-term memory systems for large language models (LLMs) have gained popularity for extending reasoning capabilities across applications. Current memory systems rely on LLM agents to organize and consolidate memory, resulting in costly, inefficient write operations. To address this limitation, we propose MemFit, a long-term memory system for conversational agents that reduces the cost and latency of memory operations. Unlike existing systems that rely on expensive LLM calls for memory construction or discard surface-level details through compression, MemFit stores each turn verbatim in an append-only store with near-instantaneous, LLM-free insertion, indexing turns with segment summaries rather than replacing them. Additionally, MemFit uses an LLM-free, multi-path retrieval strategy that combines lexical and semantic signals with cross-encoder reranking over caption- augmented episodes in both textual and multimodal settings. Empirical results on three widely used benchmarks, LoCoMo, MemGallery, and LongMemEval-S, show that MemFit achieves state-of-the-art performance while reducing memory construction time and cost several-fold, providing a scalable and efficient solution for persistent agentic memory.

We study the inverse problem of determining a connected, closed Riemannian manifold of dimension at least two from wave observations. The sources are supported in an open set $S$, and the resulting waves are observed on an open receiver set $R$, which may be at a positive distance from $S$. We prove uniqueness up to isometry within the class of manifolds that satisfy a subcritical spectral lower bound: $\|\phi|_R\|_{L^2(R)} \ge ce^{-C\lambda^\theta}\|\phi\|_{L^2(M)}$ for every eigenfunction $\phi$, where $\lambda$ is the corresponding eigenvalue, $\theta\in[0,1/2)$, and the constants $c,C>0$ do not depend on $\phi$. If a bound of this type holds also on $S$, then uniqueness holds within the class of all connected, closed manifolds. By the spectral inequality, the bound always holds with $\theta=1/2$. Uniqueness nevertheless fails at this endpoint: in every dimension $n\ge 2$, we construct a pair of nonisometric connected, closed Riemannian $n$-manifolds that produce identical wave observations.

We present an educator-guided LLM pedagogical agent for scaffolded feedback in conceptual database design. Integrated into an entity–relationship diagram (ERD) editor, the system grounds feedback in the student artifact, assignment requirements, educator-authored rubrics, and instructional resources. Its architecture separates hidden, artifact-grounded diagnosis from the workflow that controls the form and disclosure level of student-facing support. We instantiate the architecture as a four-stage workflow progressing from concept checks and guided application to low-detail feedback and localized clarification. Each feedback request creates a stateful episode linked to versioned ERD states. In a deployment spanning three ERD environments and 383 feedback episodes, 71.1% of observed target-level changes fully or partially incorporated the hidden diagnostic target, including many after Stages~1–2. Qualitative analysis showed that staged disclosure sometimes withheld inaccurate details, supported selective uptake, or allowed later recovery, though some errors still shaped revisions. Survey responses from a self-selected sample favored delayed disclosure and student agency but noted indirectness and repetition.

We study endpoint mapping properties of low-energy wave operators for Schrödinger operators $H=-\Delta+V$ on $\mathbb R^n$, $n\ge3$. In dimension four, we prove that a zero-energy resonance prevents $L^1$ boundedness, whether or not zero is also an eigenvalue, under $|V(x)|\lesssim\langle x\rangle^{-\beta}$ with $\beta>10$. For a zero-energy eigenvalue without a resonance, we obtain a complete low-energy $L^p$ classification in every dimension $n\ge3$ under $\beta>n+4$. In particular, $L^\infty$ boundedness is equivalent to the vanishing of the zeroth, first, and harmonic second moments of $V\psi$ for every zero-energy eigenfunction $\psi$. The proof identifies the finite-rank obstruction and shows that the remaining eigenvalue correction cannot cancel its critical asymptotic profiles. The same conclusions hold for the full wave operators when the corresponding high-energy bounds are available.

A 35 input/single output quantum phase neural network (QPNN) is proposed to recognize the English letters' A 'and' B 'composed 7*5 of pixels. Each node of the proposed QPNN has the input with 0 or 1. Through normalization and quantum phase encoding, each set of 35 with 0 or 1 digital inputs is encoded into a quantum superposition state described by 35 quantum phases. Then, a quantum rotation gate is used to introduce adjustable phase, and a controlled NOT gate is used to entangle the relative phase between the two inputs. The relationship between the input and output of the network is derived. This paper also derived the phase weight learning training algorithm with adaptive learning rate. The solutions with the recognition probability of 1 is designed, and the analytical expressions for all adjustable phase solutions with zero errors are not unique. The experimental performance is verified on the Qiskit platform. This paper uses phase drive to convert the probability calculation into an analytical polynomial function with adjustable rotation angles in the network, avoiding the bottleneck of generating all 2^35 complex exponential amplitudes and providing an effective way to solve the "exponential wall" problem, which provides a new implementation solution for the practical application of multi-input recognition problems.

We consider the problem of scalar-on-function regression. Most existing methods implicitly assume that the functional covariates are observed exactly, but in practice, they are often contaminated by measurement error. Our goal, therefore, is to deal with the problem of scalar-on-function regression when the regressor functions are observed with error. In this paper, we propose a simulation-extrapolation method to correct for the attenuation of estimated coefficient functions caused by the error. The method first estimates the error variance, establishes the relationship between a sequence of added error variance and the corresponding estimates of coefficient functions, and then extrapolates to the zero-error. We describe three methods to extrapolate the sequence of estimated coefficient functions. In a simulation study, we compare the performance of the simulation-extrapolation method with two pre-smoothing methods based on smoothing splines and functional principal component analysis. Next, we discuss the extension of the method in several directions, allowing for more complex noise covariance structures, multiple replications of functional predictors, generalized responses, and 2D and 3D functional predictors. Finally, we illustrate the methods by an application to diffusion tensor imaging data.

Patient-reported pain location is diagnostically decisive for some presentations and nearly uninformative for others. A companion paper argues this is not one gradient of diagnostic utility but three distinct failures of localization. This paper gives those failures their mathematics and shows they are one object: a single Bayesian generative model failing at different nodes - the likelihood, the model class, and group- or context-dependence in that same likelihood. The count is not in dispute: the first three are failures of the inference that produces a felt location, and the fourth, added here, is a failure of reporting it. Anatomical multiplexing is a non-identifiable inverse problem: a rank-deficient referral matrix sends distinct causes to one report. Delocalized amplification is a change of generative model whose dynamics are a neural-field bifurcation, with spatial extent as the order parameter. Referred and atypical displacement is group- or context-dependence in the presentation likelihood - a probability, not a loss - whose consequence sits one layer downstream, in a decision threshold varying with group prevalence and cost. Repeated observation cannot reduce recoverable information about a fixed inferential target, an exact chain-rule identity, while practical value can fall. The fourth node is the report itself. A spatial Bayesian model factorizes mislocalization into referral blur, an anatomical offset and a precision-weighted cognitive override, and reproduces phantom-limb, mirror-box and central-post-stroke reports as regimes of one equation. It is estimable under conditions the paper states as necessary or sufficient rather than assuming them; Sec. 8 gives the design that meets them. One control is reported against the paper's own interest: the transport statistic first used to measure migration scores a stationary, spreading profile as though it had travelled.

We construct a two-parameter family of single-error-correcting seven-ququart codes with transversal $\tilde{A_7}$ symmetry, realizing the finite component of a two-qubit super-golden gate set. These $((7,4,3))_4$ codes encode two logical qubits and support non-Clifford operations by applying the same gate to each physical ququart. Our method constructs invariant parent states rather than searching directly for a codespace. For $n$ systems of local dimension $q$, suppose the local symmetry is a unitary $t$-group and the permutation symmetry is $t$-transitive. Maximal mixing of all marginals on at most $t$ sites is then equivalent to linear equations in Schur–Weyl sector weights. Every feasible solution yields parents of pure distance at least $t+1$; puncturing one site produces a $q$-dimensional code of distance at least $t$ with the prescribed local symmetry acting transversally. Steane and the nonstabilizer Fake Steane code reveal why the construction leaves continuous freedom. They share the same prescribed transversal Clifford and permutation symmetries, and their eight-qubit purifications lie on a circle: error correction fixes two invariant-sector norms but leaves the relative phase free. Keeping the same permutation group and replacing Clifford symmetry by $\tilde{A_7}$ $\leq SU(4)$ gives three fixed sector norms and two free relative phases, producing the new ququart family.

In this paper, we study how to achieve one-step action generation in Robotic Foundation Models (RFMs), aiming to overcome the high inference latency of multi-step flow matching. MeanFlow provides a promising framework for this goal, yet its direct application leads to performance collapse. We discover that this stems from two distinctive dynamics exhibited in the RFM velocity field: (1) the “local acceleration" exhibits stability early on, but surges sharply towards the end of the denoising process, and (2) the spread of its magnitudes across samples widens as denoising progresses. To address these issues, we introduce Kinematic MeanFlow (K-MF), a novel one-step action policy tailored for RFMs. Specifically, grounded in a kinematic identity, K-MF decouples the time derivative term in the MeanFlow formulation into two sub-interval terms separated by an intermediate point. This decoupled formulation enables the two terms to capture early-stage and late-stage denoising dynamics, respectively, while mitigating the error amplification across the process. As a result, our K-MF empowers RFMs to achieve one-step action generation in both training from scratch and fine-tuning paradigms across diverse tasks, while outperforming multi-step flow matching in most settings. In terms of inference efficiency, K-MF reduces action-head latency of GR00T-N1.6 by 67.5%~74.4% across L40 and Jetson Orin in eager and compiled modes, yielding end-to-end latency reductions of 30.3%~54.9%. Code will be available at https://github.com/IntelChina-AI/K-MF.

Large language models can generate plausible solutions to programming assignments, making it tempting to detect their use by matching student code against a reference bank of generated solutions. Yet similar code can also arise when an assignment admits only a few natural implementations, which leaves open what a match actually shows. We investigate generated-reference matching using 29,970 student submissions from ten Python labs offered in 2021, 2023, and 2025, together with 90,000 solution attempts generated retrospectively by three frontier LLMs. We validate the generated solutions using hidden instructor tests, compare code with MOSS after excluding the starter code, and examine the exact abstract syntax tree (AST) forms of selected functions. The models usually produced correct solutions and, across most assignments, converged on similar implementations. Student submissions matched the generated references more often in later cohorts, including among submissions that passed every hidden test. On tightly specified functions, the models converged on a few exact abstract-syntax-tree forms, and the number of distinct student forms also declined across cohorts, whereas open-ended functions remained diverse in both sources. Most reported overlaps were short, making the minimum match length an important choice when reviewing students' code. Finally, we discuss how instructors can build a reference bank of generated solutions before releasing an assignment to identify tasks on which generated solutions converge, decide how much review a match warrants, and redesign tasks to elicit tests, reasoning, and intermediate work. These findings support tracking population-level changes in submitted code, while attributing AI use to an individual submission would require additional evidence about how it was produced, such as prompts, revisions, intermediate code, and student disclosures.

For many infectious diseases, behavior change is not a population-level reaction to rising case counts, but a personal one, triggered by how sick an individual feels. Yet most models fail to capture how illness severity drives this response. Multi-scale models that link the immunological scale to the epidemiological scale seem well-suited for capturing this behavior. Most multi-scale models link within-host viral load to population-level transmission rates through linking functions that are a function of viral load. However, viral load is often misaligned with symptom severity and thus possibly ill-suited for the purpose of informing behavior at a population level. We present a novel multi-scale model in which symptomatic individuals dynamically modify their transmission based on systemic symptom scores (a composite measure of muscle ache, fatigue, headache, and feverishness informed by within-host immune dynamics) while asymptomatic individuals do not change behavior. We find that stronger illness-driven behavior change delays and lowers the epidemic peak. Moreover, the resulting epidemic trajectory remains qualitatively similar to that of a standard SEIR model without behavior change, across several distinct choices of linking functions connecting the two scales. This stands in contrast to models where behavior change is driven by population-level feedback (e.g., reported cases or deaths), which can produce qualitatively distinct signatures such as plateaus, shoulders, or an elongated epidemic decline. In the case when an exponential function of symptom score is used as a linking function, we show that the standard SEIR model can closely reproduce the total number of infectious individuals at a given point in time generated through the multi-scale level.

Adversarial optimization under a shared $\ell_1$ budget requires deciding not only how much perturbation to use, but also where that limited budget should be spent. This allocation problem becomes particularly important when individual input coordinates are subject to local magnitude constraints, which restrict the extent to which perturbation can be concentrated on a small number of locations. We introduce an importance-guided allocation mechanism that uses a fixed clean-gradient prior to steer perturbation toward model-sensitive regions while leaving the feasible perturbation set unchanged. A centered allocation objective encourages perturbation at above-average importance locations and discourages unnecessary expenditure elsewhere, thereby redistributing rather than enlarging the available budget. Across ten robust model–dataset configurations under a common capacity-limited threat setting, the proposed method improves attack success over matched APGD- and PMA-based baselines by $2.52$ to $17.70$ percentage points. Allocation analysis shows that these gains are accompanied by substantially greater perturbation mass in high-importance regions without increased global $\ell_1$ consumption. Mechanism ablations further show that centered non-uniform redistribution provides part of the benefit, while model-derived importance yields an additional improvement. These results identify perturbation allocation as a distinct and practically relevant dimension of adversarial optimization under shared-budget, locally constrained threat models.

Background and Objectives: Fluorescence-labeled cellular arbors provide readouts of neuronal and microglial morphology, but fine and weakly labeled processes are prone to fragmentation and false connections that bias skeleton-based measurements. We present MorphoBranch, a fine-structure-preserving, human-reviewable workbench for morphometry of branched cellular structures. Methods: MorphoBranch combines a deterministic Morphometry Engine with an LLM-assisted Refinement Engine. The Mor- phometry Engine implements an image-to-graph workflow integrating multiscale structural evidence extraction, hysteresis segmen- tation, evidence-constrained skeleton refinement, and graph-based morphometry. The Refinement Engine maps natural-language requests to registered actions for parameter adjustment, preview execution, metric reporting, and unsupported-request handling, while image processing and quantitative computation remain deterministic and reviewable. Results: MorphoBranch was evaluated on two public neuronal axon datasets, AxonMIP and AxonStack, and the in-house Cell- Morph dataset of microglial fluorescence images. It achieved the highest Skeleton F1 and clDice and the lowest length-estimation error among the evaluated methods on all three datasets, while also achieving the highest Dice and IoU on AxonMIP and Axon- Stack. Across 150 natural-language tasks, the Refinement Engine achieved a 94.0% end-to-end success rate. Conclusions: These results demonstrate that MorphoBranch provides a reproducible, human-reviewable workflow for mor- phometric analysis of branched cellular structures. It supports fine-structure-preserving quantification across neuronal axon and microglial fluorescence images while maintaining inspectable and reproducible analysis workflows.