Papers

We consider the maximum-depth problem for mixed 2-sided orthants in R^d: each region imposes one lower bound and one upper bound on distinct coordinates, and the task is to find a point contained in as many regions as possible. We show that the corresponding decision problem is W[1]-hard when parameterized by the dimension. Our reduction from MultiColoredClique uses two coordinates per color class and only polynomially many orthants.

Benchmarks, audits, and agent protocols describe performance, permissions, and repair, but not how observed evidence should change an agent's authority during a consequential task. We propose a Runtime Assurance Contract (RAC), a policy-level formal schema binding autonomy boundaries, component eligibility, evidence state, transition policy, human-review capacity, and non-compensatory gates. Under RAC, soft metrics may inform routing, whereas a failed or unknown mandatory gate forces retry, switch, escalation, deferral, or stop; aggregate performance cannot authorize action. We define the contract, an evidence record, a permission rule, and five invariants. A deterministic failure-injection study in agentic coding evaluates 174 cases, each a distinct mechanism, under a gate conjunction, a score-only rule, and a restricted protocol baseline. At the published example weights and threshold, the score rule admits 56 of 63 block-required injections and all 24 review-required injections; tuned in hindsight, it matches the conjunction case for case. For positive weights, a positive threshold, binary risk signals, and each signal firing alone on some case, the two rules agree on every case if and only if the threshold does not exceed the smallest weight. On 124 further cases written blind by a model of another vendor, the conjunction itself admits 35 of 79 injections, and so does the tuned score rule: a missing detector is invisible to every decision rule. Eighteen hand-authored traces check version-pinned evidence and review transitions. In a prospective synthetic holdout of 24 episodes, RAC and a separately implemented full stateful baseline both match the labels of two blinded LLM judges. These studies test mechanisms on synthetic cases; they establish neither deployed safety nor cross-domain effectiveness.

Let $\mathcal{F}$ be a nonempty family of connected bipartite graphs, each with at least two edges. For a graph $G$, a proper vertex coloring of $G$ is $(2,\mathcal{F})$-avoiding if no member of $\mathcal{F}$ occurs bichromatically, and $\chi_{2,\mathcal{F}}(G)$ denotes the minimum number of colors in such a coloring. A B-coloring of $G$ is a proper edge-coloring in which every $4$-cycle is rainbow, and $q_B(G)$ denotes the minimum number of colors in a B-coloring of $G$. For a fixed connected bipartite graph $F$ with at least one edge and bipartition classes $X_F$ and $Y_F$, define $k(F)=\min\{|I|:I\subseteq X_F\text{ or }I\subseteq Y_F,F-I\text{ is a forest}\}$. Let $m\ge2$ be the minimum number of edges in a member of $\mathcal{F}$. We prove that if $k(F)\le m-2$, then every $F$-free graph $G$ of sufficiently large maximum degree $\Delta$ satisfies $\chi_{2,\mathcal{F}}(G)=O((\frac{\Delta^m}{\log\Delta})^{\frac{1}{m-1}})$, which gives a positive answer to Chuet's Problem A and C in a sharp sense, thereby extending the results of Chuet [arXiv:2603.23379] from frugal colorings to $(2,\mathcal{F})$-avoiding colorings. For B-colorings, put $k=k(F)$, $h=|V(F)|$, and $s=\min\{|X_F|,|Y_F|\}$. We prove that every $F$-free graph $G$ of sufficiently large maximum degree $\Delta$ satisfies \[ q_B(G)\le \begin{cases} \Delta+\Delta^{1-\eta}+1, & \text{if }s\le2,\\ (4h-2)(\Delta-1)+1, & \text{if }s\ge3\text{ and }k\le1,\\ C\frac{\Delta^{2-\frac{1}{k}}}{\log\Delta}, & \text{if }k\ge2, \end{cases} \] where $\eta>0$ and $C>0$ depend only on $F$. For $k\le1$, the linear order is best possible, and for $k\ge2$, the bound is nearly sharp. To prove these results, we develop a common reduction of the coloring problems to $P$-perfect matching problems in auxiliary hypergraphs and apply the forbidden-submatching theorem of Delcourt and Postle.

Quorum systems enforce strict consistency in Byzantine fault-tolerant (BFT) state machine replication: Before a value is decided, a subset of replicas (called quorum) must exchange votes for the value. In wide-area networks, the size and composition of a quorum determines the speed at which replicas can make progress and thus impacts the latency perceived by clients. A variety of quorum constructions has been proposed, e.g., threshold, weighted, grid and others, but the literature offers little in the way of an apples-to-apples comparison. Each construction typically lives inside a different protocol implementation, so the influence of quorum design on client latency cannot be easily separated from other implementation effects, especially the use of optimization techniques. In this paper, we close that gap by extending BFT-SMaRt with a modular quorum-system abstraction that allows a variety of quorum constructions to live inside the same framework and interact with the same set of optimization techniques. Via our framework, we explore the impact of different quorum system constructions on client-observed latency in wide-area BFT replication using network simulation.

Top researchers have good intuition, but do language models have as good intuition about model training as top AI researchers? To measure model intuition of LLMs and humans, we introduce the ArchitectureIQ benchmark. Each question presents a synthetic dataset and several training recipes, and the test-taker is asked to predict the recipe yielding the best test metric. Overall, we find that LLMs' model intuition is good but has four limitations: (1) The intuition is imperfect, or even sub-human in some cases. Frontier models achieve around 76% accuracy (random choice 33%) vs best human researcher (66.0%), yet remain far from perfect. For architecture-only questions, best human achieves 65% while GPT-6 Astra only has 38%. (2) The intuition is empirical, not structured, supported by the fact that more CoT compute does not lead to substantial improvement. Unlike math, we still lack a "Science of AI" language that enables structured reasoning on AI. (3) The intuition is not maximally condensed, and can be further compressed into a knoledge base. Our constructed knowledge base with only 20 items yields large gains for weak models: GPT-4o equipped with the accumulated knowledge almost matches the performance of Claude Opus 5. (4) The intuition is insensitive to dataset properties, but the best model should in general depend on data properties. This suggests that data is the real "dark matter" in AI – LLMs (so do human researchers) understand too little about data, even less than model architectures.

A precise measurement of prompt and non-prompt ${\rm D}^{\rm *+}$-meson spin alignment with respect to the helicity and production axes in proton-proton (pp) collisions at $\sqrt{s} = 13.6$ TeV is presented, using data collected with the ALICE experiment at the LHC. The spin alignment is quantified by measuring the diagonal spin density matrix element $\rho_{\rm 00}$ for ${\rm D}^{\rm *+}$ mesons at midrapidity ($|y|<0.8$). The measurement is performed in the transverse-momentum ($p_{\rm T}$) range $3<p_{\rm T}<100$ GeV/$c$ for prompt ${\rm D}^{\rm *+}$ mesons and $3<p_{\rm T}<30$ GeV/$c$ for non-prompt ${\rm D}^{\rm *+}$ mesons, respectively. The measured $\rho_{\rm 00}$ values for prompt ${\rm D}^{\rm *+}$ mesons are consistent with the unpolarised expectation of $1/3$ over the entire $p_{\rm T}$ range, indicating no evidence of spin alignment. In contrast, $\rho_{\rm 00}$ of non-prompt ${\rm D}^{\rm *+}$ mesons exhibit a deviation from $1/3$, consistent with expectations from weak beauty-hadron decays and well described by PYTHIA 8 simulations coupled with the EvtGen decay package. These results confirm and extend previous measurements, providing a precise baseline for future studies of ${\rm D}^{\rm *+}$ spin alignment in heavy-ion collisions.

We consider simulation-based Bayesian inference (SBI) for the parameters of models with intractable likelihoods but tractable forward simulation. Building on Gaussian mixtures-of-experts surrogates, as a first contribution we develop a sequential procedure to estimate the posterior, incorporating an importance sampling step that introduces a locally amortized likelihood-to-evidence ratio estimator. The local amortization allows the ratio estimator to be evaluated at new parameter–data pairs from a local (data-dependent) design, without re-estimation, which is particularly convenient for our second contribution. The second contribution is localized simulation-based calibration (SBC). Localized SBC provides a graphical diagnostic of calibration over neighborhoods that are deliberately broader than the available posterior, at low additional computational cost. Rather than repeatedly running the full inference procedure for each simulated pseudo-observation, localized SBC fits the local surrogate and ratio estimator only once for each local design, and reuses them across all pseudo-observations. This approach is particularly appealing for SBI applications. We evaluate both contributions on three case studies, including a real-data epidemiological application.

We call an $n$-tuple of $2 \times 2$ matrices a “\textit{$2 \times 2$ matrix $\Gamma_n$-contraction}” if its joint eigenvalues (joint spectrum for a commuting tuple of matrices) are contained in $\Gamma_n$. In this short note on $\Gamma_n$-contraction, we produce two families of $2 \times 2$ matrix $\Gamma_n$-contractions whose fundamental operators do not satisfy the commutativity conditions and \begin{equation*} \begin{aligned} F_iF^*_{n-j} - F_jF^*_{n-i} = F^*_{n-j}F_i - F^*_{n-i}F_j, \quad 1 \le i, j \le n-1, \end{aligned} \end{equation*} which shows that the commutativity as well as the conditions on fundamental operators mentioned above are sufficient, but not necessary, for the existence of $\Gamma_n$-isometric dilation.

Scientific abstracts mix contributions with background, motivation, and meta-language, so tools that read them as-is cannot separate what a field produces from what it discusses. We present Drift Inspector, an open-source system for measuring and exploring how a research field changes over time at the level of Atomic Contribution Claims (ACCs): decontextualized, contribution-bearing propositions an LLM extracts from each abstract before analysis. The system clusters these claims across years into an interactive map where every trend traces back to the claims and papers behind it. Applied to six years of EMNLP, it shows the field shifting away from classic NLP tasks toward LLM-era capabilities such as reasoning and multimodality – a movement that keyword or whole-abstract counts blur. The released data extend beyond EMNLP: the same pipeline has processed the full ACL Anthology (346k claims, 80k abstracts, 423 venues). Extraction is human-validated and clustering checked against an external manually constructed taxonomy.

Recent diffusion-based pipelines have achieved promising progress in image-to-3D synthesis. However, generating high-fidelity details remains challenging, especially when the input image contains rich details. Existing approaches often rely on globally encoded conditioning features, which compress spatial information and limit the model to reproduce fine-grained details. This common design often leads to a phenomenon we term detail attenuation. Moreover, improving image-to-3D synthesis quality typically requires retraining or fine-tuning large diffusion models, which can be computationally expensive and impractical for complex 3D pipelines. In this work, we present Blended Tile Conditioning for image-to-3D generation (BTC3D), a training-free inference time framework that enhances fine-grained detail preservation in image-to-3D diffusion pipelines. To alleviate detail attenuation, we first examine the image feature additivity in image-to-3D models. Based on this property, we introduce a blended tile embedding that extracts local conditioning signals from split image regional patches, allowing the diffusion model to better preserve fine-grained visual details. To integrate the global and local conditioning guidance stably, we propose a dynamic conditioning schedule that gradually increases the influence of tile-level conditioning during later low-noise stages of diffusion. Our proposed method BTC3D operates entirely at inference time and can be seamlessly integrated into existing image-to-3D diffusion pipelines. Experimental results demonstrate that the proposed approach significantly improves texture quality and visual fidelity of the base model while maintaining global structural consistency in a training-free manner.

This paper develops a cooperative integrated sensing and communication (ISAC) framework in which a multi-antenna base station (BS) simultaneously localizes a target and serves multiple communication users, a subset of which is equipped with reconfigurable intelligent surfaces (RISs). The Fisher information matrix (FIM) for target positioning is derived, explicitly characterizing its dependence on the BS beamforming coefficients, RIS phase profiles, and bistatic sensing geometry. Under the stated scaling assumptions, coherent RIS phase alignment yields a Fisher-information gain that scales quadratically with the number of RIS elements, whereas independent random phases provide a linear gain in expectation. We formulate a sensing-centric joint active and passive beamforming problem that minimizes the position error bound (PEB) subject to per-user signal-to-interference-plus-noise ratio (SINR) and transmit-power constraints. The resulting non-convex problem is addressed through an iterative successive convex approximation (SCA) procedure that solves a sequence of convex subproblems. We further develop a target localization estimator that fuses one direct time-of-arrival (ToA) measurement, one angle-of-arrival (AoA) measurement, and multiple RIS-assisted indirect ToA measurements. Under small measurement errors and asymptotically efficient first-stage ToA/AoA estimation, the estimator covariance approaches the Cramér–Rao bound (CRB) to first order. Numerical results validate the analytical scaling laws and demonstrate the localization gains enabled by cooperative RIS-equipped users.

We construct two toric manifolds of complex dimension four, which are not homotopy equivalent but have isomorphic integral cohomology rings.

Let $f$ be a primitive holomorphic cusp form of arbitrary weight, level and nebentypus. We prove $\int_0^T|L(1/2+it,f)|^6\,dt\ll_{f,\epsilon}T^{2+\epsilon}$, extending Jutila's level-one theorem. The new ingredient is a large sieve comparing the stationary phases in the Booker-Milinovich-Ng transformation at different heights. Applications include moments of order $T^{1+\epsilon}$ on every line to the right of the critical line, $\Omega(T^{4/35-\epsilon})$ simple zeros, zero density estimates, and bounds for cubic-field power sums, shifted convolution sums and the general divisor problem.

How much time is fundamentally required to extract the syndrome of a stabilizer code on hardware with a given connectivity graph? Previous bounds answer through a chosen list of checks (Delfosse, Beverland, and Tremblay, 2021) or through the global code parameters (Baspin, Fawzi, and Shayeghi, 2023). Here, we show that neither determines the interaction that a particular hardware cut requires. We therefore define the intrinsic cut cost, $\chi$, directly from the stabilizer group, independent of the generating set used to describe the code. It is computable from the check matrix alone, and equals the mutual information across the cut in the maximally mixed code state. We show that every exact extraction circuit, however adaptive, needs at least $\chi/2$ two-qubit operations across the cut, and alternatively with free local processing exactly $\chi$ shared Bell pairs. Guided by this intrinsic cut cost, we construct annular surface codes with constant-depth local extraction that give a counterexample to Corollary 1 of Delfosse, Beverland, and Tremblay. Furthermore, for a fixed quantum Tanner family, $\chi$ grows at least linearly with the size of every sufficiently small region, giving a depth guarantee at each corresponding bottleneck in every layout. This strengthens the global perspective of Baspin, Fawzi, and Shayeghi and, together with our matching square-grid construction, yields a tight space–depth tradeoff. Our results connect the interaction required by syndrome extraction to hardware communication capacity and thus provide a way to assess unavoidable circuit costs before committing to a particular layout.

This paper demonstrate that whether masking-based token pruning helps or hurts worst-group robustness can be predicted before deployment, without labels or fine-tuning. A systematic study of semantic masking across 8 spurious-correlation benchmarks shows its effect on worst-group accuracy is highly unstable: it improves accuracy by up to 82.5% relative on some datasets and degrades it by up to 100% on others. We trace this instability to spurious inversion: background patches receive higher CLIP text-similarity than the true object when the spurious attribute is background-separable, inverting the assumption every text- and attention-guided pruning method relies on. We introduce the Spurious Inversion Metric (SIM), a label-free, pre-deployment diagnostic whose sign predicts this effect with statistical significance (binomial $p=0.035$) across all 8 datasets, and remains dependable across 6 CLIP architectures with a clean foreground/background split. Naive masking is itself a major source of risk: it causes the largest average-accuracy loss of any method we evaluate, and its own per-image segmentation step is a significant runtime bottleneck. To address this, we design a batched, synchronization-free GPU segmentation routine that cuts this overhead from 3.5$\times$ to 1.75$\times$ baseline. Gating deployment by SIM's sign recovers masking's benefits while avoiding its worst failures, matching or exceeding a strong pruning baseline on 7 of 8 datasets.

Emerging non-matrix ML kernels, such as LayerNorm, GeLu, FFT or circular convolutions, demand low-latency, energy-efficient spatial accelerators beyond MatMul-centric arrays. STELLA presents a spatio-temporal elastic 16 nm coarse-grained reconfigurable array (CGRA) with a rapid configuration path, per-PE hardware loop control, and a low-latency, deeply pipelined elastic fabric with spatio-temporal data reuse. STELLA reaches up to 110 GOPS/mm2 at 850 MHz, and improves effective kernel throughput by 4.84-7.14x over baseline CGRAs.

Adapting a language model to a specialized corpus means choosing which instruction-tuning tasks to train on under a fixed budget, and testing one choice costs a fine-tuning run. Common heuristics add more source tasks or pick sources similar to the target. The first assumes transfer is never negative; the second, that it is symmetric. We show that both assumptions fail: task $A$ can help task $B$ while $B$ hurts $A$, so helpfulness is a signed property of ordered source–target pairs. We introduce the transfer map, a signed estimate of how much each source helps or hurts each held-out target. We fit the map in hundreds of fine-tuning runs on Qwen3 and Mistral models from 0.6B to 32B parameters, with all sources drawn from one corpus and no training examples from the target. The map predicts a held-out target's accuracy on unseen mixtures: recorded before those runs, its predictions have less than half the error of a mixture-agnostic baseline. The map is specific to its target and corpus but transfers across model scale: a mixture selected in advance at one size beats training on all source tasks at every other size we tested. Transfer is thus a property of the data. The map selects the tasks that help and drops the one that interferes: accuracy on the reasoning targets (causal explanation, multi-hop questions and methodological critique) rises by up to 14 percentage points over training on all source tasks.

Value steering should change an LLM's normative priorities while preserving the scenario, facts, and task constraints underlying its answer. Conventional activation edits often change both. We introduce an editable semantic-value interface on frozen residual states, with a one-way semantic-to-value pathway that grounds value recognition in context. Stop-gradient blocks feedback through this pathway; swap consistency, topic de-confounding, and decorrelation encourage selective codes. At inference, editing the value code produces a residual delta while holding the semantic code fixed. On two instruction-tuned backbones, this interface improves semantic preservation and reduces benign refusals at comparable value alignment. A matched mixing-by-gating ablation separates representation learning from selective edit activation, and dimension-matched probes establish improved code selectivity. Against validation-selected prompting on LLaMA-3.1-8B, the method achieves comparable alignment (0.750 vs. 0.748), higher BERTScore (0.938 vs. 0.923), and fewer contradictions (5.1% vs. 7.6%). Human ratings and cross-taxonomy controls provide complementary evidence for low-damage value steering.

The finite basis problem for small semirings differs from its semigroup counterpart even in order three. Recent work classifies several additive types of four-element additively idempotent semirings, including all 386 semirings whose additive reduct is a chain. We consider all four-element semirings with commutative addition, in the binary signature without named constants. Combining the existing classifications with structural reductions and polynomial normal forms, we obtain 2284 finitely based and 57 nonfinitely based isomorphism types among the 2341 types. The nonfinitely based cases consist of 45 additively idempotent semirings and twelve with nonidempotent addition. The positive arguments give explicit bases or finite constructions with specified bounds. The negative arguments use cited results, term retractions and a hypergraph obstruction. A complete catalogue records the applicable result for each representative; separate correspondence tables identify precisely the cases supplied by the earlier classifications.

We study large deviations for rare clusters in sparse systems of moving particles. In the regime \(nr_n^d\to0\) and \(\rho_{k,n}=n^kr_n^{d(k-1)}\to\infty\), we prove a large deviation principle for the empirical measure of isolated \(k\)-particle trajectory clusters. The speed is \(\rho_{k,n}\), and the rate function is the relative entropy \(h(\,\cdot\,\mid\tau_k)\), where the finite reference measure \(\tau_k\) explicitly incorporates the underlying path law. As consequences, we derive free-energy variational formulas for bounded interactions and a hard-core constraint and identify the corresponding minimizing cluster law. The normalized optimizer provides the basis for Metropolis–Hastings sampling of interacting trajectory clusters. As an application motivated by chain formation and swarming in active particle systems, we calibrate the resulting stochastic model to experimental trajectories of magnetic micromotors and find that the fitted velocity scale varies systematically with particle size and magnetic forcing.

Days Alive and Out of Hospital (DAOH) and Days Alive at Home (DAH) are increasingly used as patient-centred primary outcomes in perioperative trials, and a recent editorial has advocated DAOH for its simplicity and reliance on routinely collected data. We argue that the choice between these endpoints affects power, treatment effect estimation and missing data, and should be made at the design stage with these consequences in mind. By treating days in nursing homes or rehabilitation facilities as equivalent to days at home, DAOH assumes that being out of hospital is a valid surrogate for good recovery, an assumption that patient perspectives, including those from the NOTACS trial, call into question. Choosing DAH over DAOH therefore involves a trade-off between measuring what matters to patients more accurately and the added burden of tracking discharge destination, including a greater risk of missing data. If an intervention affects only length of stay, readmission or mortality, the two endpoints yield the same expected treatment effect, and the extra burden of tracking discharge destination brings no benefit. If it instead enables more patients to return directly home rather than to a care facility, DAOH may conceal this benefit and lose statistical power. We propose a baseline-adjusted DAH, which counts only days spent in a setting representing an escalation of care relative to the patient's baseline, as a better proxy for recovery that also accommodates hospital-at-home and virtual ward services.

Colored partitions constitute a well-established and extensively studied area of research with numerous generalizations arising from imposing restrictions on the allowed colors of the parts. In particular, restricted colored partitions have received considerable attention owing to their rich combinatorial and arithmetic properties. Motivated by these developments, we consider a colored overpartition function $\overline{\mathfrak{C}}_{m,r,s}(n)$, which enumerates the overpartitions of $n$ such that each part divisible by $m$ can be assigned any of $r$ colors, while each part not divisible by $m$ may be assigned any of $s$ colors. We investigate an asymptotic formula and several arithmetic properties of this function, including its density properties.

Synthetic dialogues generated by LLM pipelines now serve as complete conversational-recommendation benchmarks: an LLM listener talks to an LLM recommender, and the track logged next in the conversation becomes the official label for each turn. These policy-selected labels make large-scale evaluation reproducible, but they are proxies for what the simulated user asked. We audit the one place where label and request are directly comparable: turns where the user asks for an exact song by name. In the RecSys Challenge 2026 TalkPlay benchmark, using visible dialogue and catalog metadata alone, we find that the official label contradicts the user's exact-song request in half of the audited development turns. This matters beyond one benchmark: naming the desired item is the dominant intent in real music search, where deployed systems avoid substituting an alternative for an exactly named item, on the premise that it costs satisfaction. A small training-time supplement closes most of the gap: adding catalog-resolved request-satisfying targets to a small fraction of training turns yields a 53.3% relative gain in nDCG@20 on the 43 conflict turns while leaving the official metric intact, verified against a matched control that detects the same requests but trains only on official labels.

Based on second-order inertial dynamics, we develop two accelerated Lagrange multiplier (LM)-based optimization methods: Linear LM-based second-order flow method and linear LM-based variable and operator splitting method for unconstrained non-convex optimization problems. Relaxed and adaptive variants of both methods are proposed to prevent degeneration of the Lagrange multiplier and to enable automatic adjustment of the time step size. Within the novel second-order LM-based optimization framework, the associated pseudo-energy functional is shown to dissipate monotonically along the iterative trajectory, and convergence of the generated iterates to stationary points is rigorously established. Extensive numerical experiments on function optimization and partial differential equation benchmark problems, including applications as optimizers for physics-informed neural networks (PINNs) and deep operator networks (DeepONets), demonstrate that the proposed methods effectively escape local minima, achieve high accuracy, and exhibit strong robustness with respect to the choice of the initial learning rate during neural network training.

Plane gravitational waves play a role both in relativistic and in non-relativisitic physics, highlighted by two different mass parameters. The large-distance behaviour of a sandwich gravitational wave can be approximated by a continuous but not necessarily smooth profile, providing us with a simplified description of particle motion. Our model is consistent with the Carroll symmetry. Our strategy is illustrated by the Pöschl-Teller profile. The limitations of the sandwich approximation are discussed.

Here we demonstrate quantum–classical break-even for chemical dynamics applications, reaching useful accuracy with a quantum-hardware wall-clock time comparable to that estimated for practical classical computing resources. To this end, we develop a workflow for problems that are hardware-native but classically hard. We target large ring-shaped molecules that map naturally onto quantum hardware with neighboring connections. Exploiting the spatial locality of interactions, we compress short-time evolution within a small region using tensor networks. We then use rotational symmetry to replicate the compressed circuit in space and time, enabling long-time dynamics of large systems with practical classical processing times. We apply the workflow to conjugated macrocycles with rotational periodicity, 12-qubit benzene and 120-qubit [60]annulene, by embedding their electronic structure onto loops of IBM quantum processors. For the annulene, we compared the quantum-processing-unit (QPU) wall-clock times for hole-doped nonequilibrium dynamics to the estimated wall-clock time of two state-of-the-art classical methods, Majorana propagation and matrix-product-state time evolution based on the time-dependent variational principle. We show a situation in which a hybrid quantum–classical strategy is comparable to a classical-only counterpart in terms of both accuracy and computational time for a conventional high-performance computing system under ideal parallelization. These results show that quantum hardware can become a realistic computational option for chemical simulations for practitioners in chemical science and engineering.

On-policy distillation (OPD) has demonstrated two important capabilities in language models: compressing large teachers into smaller students and merging expert models into a single model. Existing diffusion OPD, however, mostly focus on the latter, with teachers and students sharing the same backbone and scale. We investigate large-to-small diffusion opd from large teachers to a small student and find that the standard recipe fails. To find the underlying cause, we propose Fixed-State KL, an effective and fair way to measure the distribution gap between student and teacher during OPD training for diffusion models. We are the first to clarify why large-to-small OPD is challenging for diffusion models: a smaller student struggles to perfectly match the distribution of a larger teacher, while classifier-free guidance can accumulate and amplify the distributional discrepancies between the student's conditional and unconditional branches and those of the teacher. To solve this problem, we propose GFD-OPD, a simple yet effective method that reduces the student-teacher gap while avoiding the error amplification of the CFG composition. Across numerous experiments, GFD outperforms previous baselines in both training efficiency and final performance, achieving state-of-the-art results on all benchmarks.

We show that the light-ray algebra of unitary supersymmetric Lorentzian conformal field theories contains a universal supersymmetric generalized $\mathfrak{bms}$ algebra. We construct parallel light-ray operators using the supersymmetry current, $R-$symmetry current and the stress tensor supported on a co-dimension one null hypersurface and determine their graded commutators in $d=3,4$ to verify this claim. Our construction only relies on global supersymmetry and fundamental physical principles such as microcausality, unitarity and Lorentz invariance and thus applies to strongly coupled theories. As a corollary, we show that the averaged null energy operator is a positive operator using the local supersymmetry algebra on the null sheet, providing a supersymmetric derivation of the averaged null energy condition (ANEC). Finally, we verify our results in free massless supersymmetric theories and present arguments to support that the local supersymmetry continues to ensure the ANEC is satisfied even in general supersymmetric quantum field theories.

As quantum networks move toward multi-user architectures, a central question is how the performance of entanglement distribution scales with the numbers of network nodes $N$ and active users $M$, yet existing studies use different objectives, resource assumptions, and normalizations. To account for delivery distance under multi-user demand, we define entanglement transport, which weights each delivered Bell pair meeting a fixed fidelity threshold by its endpoint separation. In a two-layer framework, (i) a resource layer measures capacity per activation slot under different aggregation regimes, and (ii) a physical layer measures service per second under implementation constraints. For matched implementations with a bounded number of activation slots per second, asymptotic-aggregation capacity sets a service ceiling, whereas achievable service must be established separately. On expanding honeycomb lattices with one random session, fixed-window entanglement-percolation constructions achieve $\Theta(N^{-1})$ entanglement transport per activation slot below their thresholds and $\Theta(\sqrt N)$ above them. For homogeneous chains, asymptotic aggregation exceeds fixed-window capacity by a factor $\Theta(N^2/M)$, and in spatial expansion a noisy physical implementation achieves $\Theta(M/N)$ service per second. These examples show that entanglement-transport scaling depends not only on connectivity but also on resource aggregation and physical implementation. The framework provides a common basis on which quantum-network scaling laws can be established.

Dwarf and low-mass galaxies are highly sensitive to stellar feedback and gas flows, yet their spatially resolved gas-phase metallicity remains poorly characterised. We measure radial gas-phase metallicity gradients for 382 star-forming low-mass galaxies from MaNGA DR17, spanning $7.2 \leq \log(M_{\star}/M_{\odot}) \leq 9.1$ and $z \leq 0.05$. Spectra are stacked in elliptical radial bins and metallicity is derived from the N2 index. After removing the primary dependence on local stellar mass surface density, we define a residual metallicity $\Delta Z$ to isolate secondary dependencies. Gradients are flat on average (median slope $-0.007$\,dex\,$R_e^{-1}$). The residual gradients correlate most strongly with light concentration ($r = 0.304$, $r^2 = 0.092$, $p < 0.001$) and gas velocity dispersion ($r = 0.204$, $p < 0.001$), though both account for at most 9 per cent of the variance. Large-scale environment shows no significant correlation. The concentration result is robust to measurement uncertainties and persists with a stricter resolution cut, suggesting internal structural properties shape the radial metal distribution in low-mass galaxies.