论文库

When humans hand each other objects, they incorporate both geometric and semantic information into this process. For example, passing a knife with the handle towards the recipient, rather than the blade, is both more ergonomic and safer. Recent state-of-the-art methods for task-oriented robot-human handovers have progressed from modeling object geometry to incorporating object affordances. However, they often forgo predicting the explicit, task-specific hand poses a human selects to utilize an object. Since many objects support multiple interaction modalities, e.g., a claw hammer used to strike or pull nails, this variability must be modeled to achieve robust task-oriented handovers. To tackle this, we propose a novel approach, GENESIS-Handover (GENErative HypotheSIS), which leverages VLM image generation to produce a variety of task-specific hand-object interaction hypotheses. These hypotheses are matched in real time to the observed human hand pose, enabling inference of the most suitable handover configuration. By leveraging VLMs as priors of plausible hand-object interactions, the method produces task-conditioned handover strategies for previously unseen object-task pairs. We evaluate the standalone interaction proposal module before deploying the full system on a mobile manipulator. In a user study with 12 participants across five task-object pairs, 83.3% perceived our method to have better task understanding than the previous state of the art.

Static application security testing (SAST) tools such as CodeQL serve as security gates in continuous integration and code review, and a gate is also read in reverse. When a previously reported alert disappears after a change, developers and the tools around them often take the vulnerability to be gone. That inference holds only if the analyzer is robust to changes that keep the vulnerability. We test this robustness for CodeQL's JavaScript taint tracking with natural, vulnerability-preserving rewrites. A large language model, given only the alert CodeQL exposes and no knowledge of its implementation, rewrites one function along a reported flow as a plausible maintenance change, and we verify by manual audit and controlled experiments that the flaw survives and the alert does not. On 172 flows that CodeQL detects in real npm vulnerabilities, three rewrites per flow silence 26.2% of them, at about USD 0.51 per silenced flow. The same generator, asked only to refactor the same functions with no knowledge of the alert, silences none of them. Every silencing rewrite traces to one of seven gaps at the library-modeling boundary, and 64 of the 70 rewrites cross a gap inside an API CodeQL already models. The largest is the WHATWG URL constructor, which propagates taint to a URL's query components but not to its pathname. The boundary is not specific to CodeQL. 74% of the silencing rewrites also silence an independently built taint analyzer, and 53% escape three analyzers combined. A disappearing alert after an ordinary-looking change is therefore weak evidence that a vulnerability has been fixed, the gaps point SAST developers to narrow and concrete repair targets, and robustness to natural rewrites belongs alongside detection rate in how taint analyzers are evaluated.

DNA-encoded libraries (DELs) are commonly used to find small-molecule binders to a target protein. But the hard part is no longer finding a hit. It is building a machine-learning model good enough to guide the medicinal chemistry that follows, one that can resolve the shift in affinity when a chemist changes a single atom. Such a model needs two things from its training data: dense sampling of the small patch of chemical space around the hit, and an adequate number of genuinely tight binders within it. A diverse naive DEL gives neither. It samples an enormous space, so any one neighborhood is covered far too thinly, and the tightest binders are rare. A model trained on that naive DEL data must extrapolate to binders tighter than any it has seen, and that is where models can fail. One way to get training data with both properties is a second-generation DEL built around the hit. We ask how complex that library must be. We treat binding as a sum of small, independent interactions, called pixels on the target and antipixels on the ligand. From the spread of affinities measured in the naive library, with no structural information and nothing yet synthesized, we estimate the complexity such a library must have to supply that dense sampling and those tight binders. The method does two useful things. It turns the number of tight binders a model needs into the number of compounds a library must contain. And it estimates the expected yield of a randomly designed second-generation DEL, so the gain from designing the library with machine learning or active learning can be measured against a baseline, without the expense of a control library as large as the experiment itself.

Evidence accumulation models (EAMs) describe decisions as noisy evidence accumulating toward a decision boundary. Although widely studied in laboratory tasks, their application to dynamic, real-world behavior has received less attention. Traffic offers a setting where evidence changes as situations unfold, response timing matters, and decisions occur within ongoing control and interaction. This semi-systematic review synthesizes 28 publications (2014-2026) that model traffic decisions and control actions with EAMs. We organize the literature by modeling level (discrete decisions versus continuous action control) and implementation (stand-alone accumulators versus models embedded in broader perception-action or interaction frameworks). We examine how model components are implemented and evaluated across tasks. Overall, EAMs account for a range of dynamic road-user behaviors. Several comparisons suggest that response timing is better described by accumulating evidence as situations develop than by fixed delays or thresholds on momentary evidence, although accumulation models do not consistently outperform other kinds of models. Whether decreasing boundaries improve the fit differs between studies, and the models offer testable accounts of how decision-relevant information is processed. For cognitive scientists, traffic offers a testbed for studying how evidence from changing environments shapes decision-making. For applied modelers, the findings show how EAM components can be adapted to different traffic tasks, and why the choice of components needs to be evaluated for each task. However, most publications that estimated parameters pooled across participants, and most were not tested beyond the data they were fitted to, highlighting the value of participant-level or hierarchical estimation and of testing on new data.

The signature transform is a principled feature map for continuous-time paths, valued for its uniqueness and universality. Recovering a path from its truncated signature is, however, structurally ill-posed, because the truncated signature map is not injective. We therefore reframe truncated signature inversion as a probabilistic problem---learning the conditional distribution of a path given its truncated signature---and adopt a signature-conditioned flow matching model as a practical estimator. This formulation elucidates the fundamental difficulty of inversion: posterior-sampling reconstruction error quantifies the irreducible uncertainty remaining after conditioning on a statistic. For Gaussian process families, we derive this error from the tractable posterior under linear-statistics conditioning, obtaining a closed form for log-GBM and numerically tractable formulas for drifted fBM and OU. This oracle error upper-bounds the corresponding error under truncated-signature conditioning, since truncated signatures are more informative than linear statistics. The experimental results are consistent with the theoretical analysis: the learned sampler under linear-statistics conditioning closely matches the theoretically derived posterior in both reconstruction error and distribution, while reconstruction errors decrease under richer truncated-signature conditioning. Moreover, generated paths faithfully recover the conditioning signature and preserve key distributional and temporal structures. Together, these results establish a well-posed probabilistic framework for truncated-signature inversion, with applicability demonstrated on financial data beyond the process families covered by theory. Code is available at https://github.com/Junoh-Kang/ProbSigInv.

We introduce a mathematical framework for studying contraction of the distributional Bellman operator under Maximum Mean Discrepancy (MMD). We show that contraction reduces to conditional positive definiteness of an associated kernel constructed from the base kernel, which we call the Bellman difference kernel. Focusing on radial kernels, we use Schoenberg's theorem to connect this condition to Bernstein functions. This gives contraction criteria that hold in any dimension and can be verified analytically or tested numerically. Our framework unifies known results for the Gaussian RBF, power-distance, and multiquadric kernels. A numerical search identifies 11 new kernels that induce contraction. We also propose chaining, a general construction that produces kernels inducing contraction with any prescribed factor. To reconcile theory with practice, we show that contraction on a set only requires conditional positive definiteness on that set. This explains why the RBF kernel can remain a valid choice when rewards are bounded, despite its lack of global contraction. Finally, we identify a fundamental tradeoff between contraction rate and discriminative power, where kernels that contract faster tend to discriminate less between distributions. Code is available at https://github.com/BaptisteDebes/BellmanContractionMMD

Enforcement defenses against prompt injection, such as CaMeL, Progent, IPIGuard and DRIFT, ask the agent's own model to write a security artifact (a program, a privilege policy, a dependency graph or a constraint list) that code then enforces. We ask whether small language model (SLM) agents can be trusted to write these artifacts. Across 15 open-weight models, 13 of them between 1.7B and 12B parameters, they cannot: the policies, graphs and constraints that the agent's model writes admit attacks on 14 (Progent), 13 (IPIGuard) and 11 (DRIFT) of the models, and CaMeL's programs fail to run in up to 68% of episodes. We present CapLite, which derives its policy from the tool schemas, commits the write calls before any tool result is read, and admits a write only if every argument that decides whom or what it affects is stated in the user's request. This guarantees write integrity: injected text cannot add a write, and no write reaches a recipient, amount, object or destination that the user did not state. No attack succeeds in 13,935 attacked AgentDojo episodes. CapLite reaches utility under attack comparable to CaMeL's (29.0% vs. 27.2%) with fewer tokens than every other enforcement defense and than the undefended agent: 17.1k per episode, 18% of CaMeL's, and the highest utility per token of all enforcement defenses.

As autonomous AI agents increasingly interact in persistent populations, their collective capabilities and alignment depend not only on individual models, but also on the institutions governing their interactions. Yet AI research has largely focused on individual agents. In this Perspective, we argue that institutional design must become a distinct focus of AI research. Drawing on institutional political economy, social epistemology and organizational theory, we develop seven principles for designing AI institutions that sustain cooperation, enable reliable collective judgments and remain accountable to legitimate human purposes. These address reciprocal contributions, accountable authority, functional differentiation, epistemic pluralism, criticism and correction, human direction, and institutional adaptation. We illustrate their application through machine science, where autonomous research challenges institutions developed around human capabilities and constraints. We then propose an experimental research programme in which matched populations of AI agents operate under alternative institutional arrangements, and describe MoltJournal, an API-first journal for autonomous research agents, as an illustration of how such institutions can be made programmable and observable. Together, these contributions establish institutional design as a research agenda concerned not only with aligning autonomous agents, but with understanding how institutions can transform them into capable, reliable and accountable collectives.

Embodied foundation models hinge on diverse robot interaction data, which are costly to collect. Egocentric human videos offer a complementary source of experience, and latent action models (LAMs) provide a popular means of leveraging these videos for policy supervision. Compared with world-action models (WAMs), LAMs can offer greater training efficiency and potentially lower inference costs by avoiding explicit pixel prediction during policy learning and execution. However, most existing LAMs learn latent actions solely through video reconstruction, which can cause their representations to cluster by visual appearance rather than by the underlying actions. As a result, they struggle to align human actions with their robot counterparts. To bridge the gap, we introduce SALA, which emphasizes establishing the semantic correspondence for latent actions. Without requiring paired data, we use the language descriptions of unpaired human and robot clips to estimate semantic similarities and supervise cross-embodiment correspondence. Specifically, we translate these similarities into soft neighborhood structures that guide human latent actions toward semantically related robot latents, while robot action reconstruction grounds the latent space in control. The resulting shared latent action space provides a unified supervision signal for training an embodied foundation model on both human and robot videos. Notably, latent action supervision enables scaling along two axes: (1) pre-training on increasing amounts of general human video consistently reduces robot action prediction error on downstream tasks, and (2) post-training on increasing amounts of task-specific human video substantially improves real-world robot task success rates.

Pore-space evolution and fluid displacement influence how rocks transmit and store fluids, affecting geological carbon storage, underground hydrogen storage and groundwater contaminant transport. A single time-resolved micro-CT experiment may produce tens to hundreds of three-dimensional images, and segmentation is required to quantify physical properties and run image-based simulations. Traditional methods such as thresholding perform poorly on noisy, low-contrast images and need manual adjustment per image; 3D deep learning networks segment each image without temporal information, so their errors vary between images and could be mistaken for physical change. Here we introduce a transformer-based 4D segmentation framework combining spatial and temporal information. It uses sequential images to resolve ambiguous phase assignments and separate physical change from segmentation error. The method is tested on images of mineral dissolution: it measures pore-volume growth and rock boundaries more accurately than an equivalent 3D model that ignores temporal information. It also captures the increase in permeability verified against reference segmentations and flow simulations. The advantage transfers to another dissolution sample without retraining. On two other datasets of fluid displacement, it reduces spurious changes in predicted rock and porosity fluctuations and accurately identifies filling events. Spatiotemporal segmentation turns sequences of hundreds of three-dimensional images into more reliable quantitative measurements of pore-space evolution and fluid displacement without per-image parameter adjustment to support the analysis of time-resolved CT imaging.

Paper published at NeurIPS 2026. For data-driven PDE surrogates, generalizing across a wide range of initial conditions and physical coefficients requires a training set of numerical simulations that consistently covers the resulting dynamics. Yet uniform sampling of these configuration parameters can leave challenging regimes insufficiently covered, leading to high prediction errors and large error variance in the trained surrogate. To address this sampling imbalance, we introduce the Online Generative Active Sampling (OGAS), which continuously adapts configuration sampling during surrogate training, as an alternative to classical round-based active learning. As the surrogate learns from a stream of simulation data generated online, OGAS trains a fast diffusion model in parallel to sample configuration parameters conditioned on a training difficulty signal, such as loss or uncertainty. This conditional sampler enables OGAS to parametrize subsequent simulations to generate challenging training data by prioritizing configurations with high difficulty values. We evaluate OGAS on 2D PDEs with distinct challenging dynamics-Kuramoto--Sivashinsky, Navier--Stokes, and Gray--Scott-using multiple surrogate architectures and up to 308 configuration parameters. Across the tested PDEs and architectures, OGAS consistently reduces the 99th-percentile prediction error and overall error dispersion compared with uniform sampling, while wall-clock overhead is negligible. Code and experiments are available at https://github.com/melissa-sa/ogas.

We study contextual bandits with $K$ actions, a class of $N$ policies, and i.i.d.\ contexts from a distribution that the learner can sample. For any sequence of loss functions fixed before play, we give an oracle-efficient algorithm with expected regret $O(\sqrt{KT\log(2N)}\,\log(2KT))$ against the best policy in hindsight on the realized contexts. This improves the $T^{2/3}$ dependence on the horizon in prior oracle-efficient bounds and achieves the optimal regret rate in this setting up to logarithmic factors. Our algorithm perturbs policy scores using a Poisson sample of auxiliary contexts. At each sampled context, the perturbation assigned to an action is a Laplace noise draw multiplied by a scale specific to that context and action. We show that choosing these scales to minimize a convex objective determined by past observations yields the stated regret bound. We further approximate the minimizing scales by projected stochastic gradient descent using only an exact offline cost-sensitive optimization oracle that returns the optimum value. Since the table of scales can be large, the algorithm represents the SGD iterates implicitly by storing the random draws that determine each update and evaluates the scales only at queried contexts. Both the oracle complexity and the remaining resource bounds are polynomial in $T$, $K$, and $\log(2N)$.

Memory decides which LLMs an edge device runs, and how fast: capacity on a $4$ GB Jetson Nano, bandwidth on an $8$ GB Jetson Orin Nano Super. Deployed formats (GPTQ, AWQ, $\texttt{llama.cpp}$'s K- and I-quants) spend equal bits on every integer code of a matrix, although the codes are far from uniform. $\texttt{LazyWeights}$ keeps these formats but tunes GPTQ so the codes compress, Huffman-codes them losslessly in interleaved lanes, and decodes them on demand in a small unit between DRAM and L2, so GPU kernels run unchanged and DRAM moves only coded bytes. At $1\%$ and $3.5\%$ perplexity budgets, coded Phi-3-mini, Mistral-7B and SmolLM-1.7B are $5.7-31.4\%$ smaller than the best $\texttt{llama.cpp}$ format. A bit-exact $0.21$ mm$^2$ ASIC decoder unit ($1$ GHz, $7$ nm) cuts Orin latency and energy per token by up to $36\%$ and $38\%$ ($54\%$ and $44\%$ with prompt-lookup speculation). On the Nano, $\texttt{LazyWeights}$ needs $58-64\%$ less time and $53-57\%$ less energy per token than CPU $\texttt{llama.cpp}$, and runs a streamed Mistral-7B 6.6$\times$ faster at $3.89$ bits per weight. A Huffman decoder beats an rANS decoder in rate and area at equal throughput; the ASIC decoder unit runs bit-exact on a Zynq UltraScale+ FPGA (AMD ZCU102).

In NLP, models trained via Empirical Risk Minimization (ERM) often exploit spurious correlations, such as dialectal markers, at the expense of robust linguistic features, thereby degrading their out-of-distribution (OOD) generalization. Invariant Risk Minimization (IRM) aims to suppress such shortcuts by identifying predictive relationships that remain stable across environments, but its practical utility for NLP remains debated. We use a causality-driven framework to evaluate IRM under three shortcut-generating mechanisms: label-generated features, confounding, and selection bias. Our results show that IRM successfully rejects shortcuts across the evaluated causal and anti-causal structures, including under limited environmental variation and increasing label noise. We then extend our analysis beyond controlled synthetic artifacts to two less constrained settings: gender-profession associations in Bias in Bios and semantically loaded actor-name shortcuts in ActorCorr. These experiments show that invariant learning can mitigate more realistic textual shortcuts, while semantically meaningful features require careful calibration of the invariance penalty.

We develop a mathematical abstraction for uniformly exploiting symmetries in programming and reasoning based on the mathematical proof principle `without-loss-of-generality' (w.l.o.g.). Given: group actions that describe symmetries of the input and the output; and a canoniser, which captures how arbitrary inputs get sent to canonical ones, our construction extends any function from the canonical inputs to the whole domain. We characterise the conditions on the canonisers which guarantee that the construction gives the unique equivariant (i.e. symmetry-preserving) extension. Furthermore, we adapt the construction to a dependently-typed setting, where it captures traditional symmetric w.l.o.g. arguments via the Curry-Howard correspondence. Finally, we extend the construction to algebraic data types, where the symmetries of the constructors both uniquely determine the symmetries of the data type and pave the way for efficient execution.

Combining vision-language models (VLMs) with parameter-efficient fine-tuning (PEFT) offers a promising approach to few-shot whole-slide image (WSI) classification in computational pathology. This study evaluates this approach for predicting ethanol drinking status from nonhuman primate liver histology. Adapters are evaluated against non-PEFT configurations on a subset of the Monkey Alcohol Tissue Research Resource (MATRR). All evaluated PEFT configurations achieve higher mean balanced accuracy than the non-PEFT baselines. For example, in the four-shot setting, attention-based multiple instance learning (ABMIL) alone achieves 43.5% mean balanced accuracy, compared with 78.1% after adding an adapter. Qualitative attention analysis shows that textual pathway adaptations produce localized attention heatmaps, while visual pathway adaptations produce more diffuse ones. A pathologist's review identified similar histological features across adaptations.

Quantifying disease-associated renal tubular remodeling is challenging because tubular appearances are heterogeneous. We introduce TuMAP (Tubular Morphotype-Aware Pathomics), a pipeline combining (1) dual-view foundation-model representations and provisional morphotype annotation, (2) compositional and within-morphotype pathomics, and (3) patient-level statistical inference. From 456,158 tubules in 358 PAS-stained kidney biopsy slides, TuMAP identified eight image-based morphotypes, seven of which showed stable subsampling assignments. Across 219 patients in six diagnostic groups, diabetic nephropathy (DN) and tubulointerstitial nephritis/acute tubular injury (TIN/ATI) each differed from minimal change disease in seven morphotype proportions, while IgA nephropathy differed in six. DN showed reduced nuclear solidity and increased eccentricity across all eight morphotypes. Area-normalized nuclear density exposed additional within-morphotype differences. TuMAP provides a framework for separating changes in tubular morphotype composition from phenotypic differences within morphotypes.

Merging independently fine-tuned checkpoints into a single multi-task model requires no data, no labels, and no gradient steps. In practice the task vectors of different tasks interfere, and the standard remedies--averaging, magnitude pruning, sign election--operate entry by entry, ignoring that each delta induces a low-dimensional update on the hidden features it acts upon. We recast merging as a subspace interference problem. For every two-dimensional weight tensor we take the top-r right-singular subspace of each task's delta--the input directions along which that task moves the representation most--and remove from each task vector the component lying in the union of the other tasks' subspaces. The resulting Joint Null-Space Projection (JNSP) is closed-form and order-independent, needs no forward pass, and is defined entirely by the checkpoint set. On eight CLIP image-classification tasks JNSP merges ViT-B/16 experts to 89.45% average accuracy without using a single image, 11.88 points above the strongest parameter-wise training-free baseline (TIES-Merging, 77.57%), 14.22 above Task Arithmetic, and 0.98 above WUDI's data-free consensus merge (88.47%). The gain holds across three backbones on the independently trained TALL release (+10.8, +12.6, +9.1 over TIES for ViT-B/16, ViT-B/32, ViT-L/14) and replicates there at 89.20%; controls confirm the source--a single shared subspace instead of per-task ones loses 2.9 points, and a sequential variant collapses to 69.6%. On RoBERTa-large the same projection underperforms plain task arithmetic, which we report rather than hide. Code is available at https://github.com.

An unconditionally secure $(t, s, v)$-all-or-nothing transform (AONT) is a bijection from $s$ input elements to $s$ output elements over an alphabet of size $v$, such that given any $s - t$ output elements, no information can be obtained about any $t$ input elements. While linear AONTs have been extensively studied, comparatively little is known about the existence of non-linear AONTs. In this paper, we investigate unconditionally secure cryptographically non-linear $(t, s, v)$-AONTs, distinguishing between non-additivity, non-affinity, and cryptographic non-linearity. To achieve this, we first present a computational method for enumerating general AONTs focusing on the case $(t, 3, 3)$-AONTs for $t = 1 and 2$, establishing the existence of non-additive AONTs. We then introduce a general construction of non-additive AONTs from linear ones using coordinate-wise symbol permutations. We measure the non-linearity of the AONT instances generated from this construction, using differential $Δ_F$-uniformity and Walsh magnitude. Finally, we establish theoretical conditions under which $Δ_F$-uniformity is reduced and present statistical results examining the non-linearity of AONTs obtained through our construction.

P300-based brain-computer interfaces (BCIs) enable communication by decoding stimulus-evoked EEG responses into character selections without requiring voluntary motor output. Reliable P300 detection remains difficult because of low response amplitude and signal-to-noise ratio, while practical spellers require accurate character selection with few stimulus repetitions and low computational cost. We propose a compact deep learning architecture that combines xDAWN spatial filtering for event-related potential enhancement and dimensionality reduction, a fixed frequency-selective filter bank, and convolutional feature learning for spatio-temporal EEG decoding. Depthwise-separable convolutions and progressive average pooling reduce computational complexity while retaining discriminative temporal information. The resulting target/non-target classifier contains only 5,100 trainable parameters. We evaluated the model on two public benchmarks, BCI Competition II Dataset IIb and BCI Competition III Dataset II, and compared it with established P300 decoding methods across different numbers of stimulus repetitions. The model achieved an information transfer rate of 17.8 bits/min with two repetitions on BCI Competition III and 40.5 bits/min on BCI Competition II, demonstrating a favorable accuracy-speed trade-off. These results suggest that combining domain-informed spatial and spectral priors with compact neural networks can support efficient P300 decoding for EEG-based BCI communication.

In Non-Uniform $k$-Center, we are given a metric space $(V,d)$ and $L$ levels of balls with prescribed radii $r_1 \ge \ldots \ge r_L$ and nonnegative integer budgets $k_1,\ldots,k_L$. The goal is to place at most $k_\ell$ centers of each level $\ell \in [L]$ in $V$ and minimize the dilation $λ$ such that the corresponding balls of radii $λr_1,\dots,λr_L$ cover $V$. While this problem admits no subexponential approximation factor unless P=NP, the picture changes drastically when additional centers are allowed. An $(α,β)$-bicriteria approximation to Smooth Non-Uniform $k$-Center covers $V$ with dilation $βλ_{\text{opt}}$ such that, for each $\ell \in [L]$, the number of centers of the first $\ell$ levels is at most $α$ times their total budget. For every fixed $\varepsilon>0$, we give a randomized $(1+\varepsilon,8+\varepsilon)$-approximation and a deterministic $(O_{\varepsilon}(1),2+\varepsilon)$-approximation, both running in polynomial time. Even for classical $k$-Center, obtaining an $(O(1),2-δ)$-approximation is NP-hard for every fixed $δ\in (0,1]$. Thus, our second result almost settles the optimal dilation when a constant-factor violation of the budget is allowed, unless P=NP. We combine a new primal-dual enumeration procedure with a simple bucketing argument that yields a new trade-off between the two approximation factors. It remains open whether there is a polynomial-time $(1+\varepsilon,2+\varepsilon)$-bicriteria approximation, which would almost match the classical $2$-approximation for $k$-Center.

The rapid growth of artificial intelligence is driving demand for new data centers at large scale and on short deployment timelines. To power them, firms connect to the commercial grid or contract generation directly, but interconnection and gas-turbine delays now incentivize on-site generation. We compare grid-connected and off-grid data center energy infrastructure (AC- and DC-coupled solar-plus-storage, islanded natural gas, and grid interconnection) across the continental U.S. Using an open-source technoeconomic framework built on empirically calibrated AI training loads, climate-driven cooling, stage-by-stage power conversion, and physics-based battery degradation, we model designs providing equivalent energy services. Our central contribution is a speed-adjusted cost of energy that values the compute unlocked by faster construction, making options with different deployment timelines directly comparable on one physically consistent basis. We find DC coupling reduces lifetime energy costs by 18% ($0.09/ kWh) on average versus AC counterparts by eliminating inverters and improving efficiency. Though rarely cheapest on unadjusted energy cost, DC-coupled microgrids' faster deployment makes them the lowest-cost modeled option in 85% of U.S. locations once foregone compute is valued, or 55% under accelerated natural gas timelines. For AI infrastructure, deployment speed rather than unit energy cost determines the economically preferred path to power.

Large language model (LLM) inference at datacenter scale is fundamentally bottlenecked by memory capacity and bandwidth limits. Because model weights are read-only and identical across instances, conventional architectures that replicate these parameters across the local DRAM of tens of thousands of accelerators incur massive capital expenditures, significant static power dissipation, and redundant memory-fetch energy. A recent proposal to use optical fiber loops as delay-line memory alleviates these issues, but its recirculating loop must hold the whole model: loop length grows with model size, slack is fixed when the ring is built, and the layouts that serve modern Mixture of Experts (MoE) models efficiently do not fit in it. We propose Optical Broadcast Memory (OBM), an open-ended hierarchical Tap-and-Amplify fiber architecture that generalizes and replaces the optical ring paradigm. A centralized memory server continuously broadcasts weights to distributed inference nodes that passively tap the data stream using receive-only co-packaged optics, delivering parameters directly to compute units without intermediate buffering. To overcome the limitations of continuous weight streaming, we use dynamically-sized slack to accommodate large and variable context lengths. We show how removing the restrictions of the optical ring allows OBM to efficiently serve MoE models. Evaluated at a datacenter scale of 80,000 accelerators, OBM eliminates petabytes of HBM while being competitive or better in throughput and energy per token than an HBM-based system.

We present new upper and lower bounds on the minimum distance of certain generalized bicycle (GB) codes that go beyond the capabilities of techniques based on classical codes. These bounds are shown to capture the true minimum distance for specific subclasses within the GB family. In particular, we demonstrate the existence of two highly degenerate GB code families with parameters $[\![d^2+1,2,d]\!]_2$ for odd $d \geq 3$ and $[\![d^2,2,d]\!]_2$ for even $d \geq 4$, both having qubit connectivity of four, similar to that of surface codes. For the odd-distance family, we analyze the structure of low-weight logical Pauli operators and establish the existence of a fault-tolerant logical CNOT gate between the two logical qubits, achievable via a simple relabeling of data qubits. Next, using nearly self-inverse polynomials, we extend our results to construct an infinite class of connectivity-six GB codes with an improved rate. For GB codes based on these polynomials, we further present a technique to realize transversal logical CNOT gates via data-qubit permutations. We further construct syndrome extraction schemes for our connectivity four GB families that avoid reductions in minimum distance due to circuit-level faults propagating from check qubits to data qubits. Finally, we numerically evaluate the logical error rate performance of the mentioned connectivity four families of GB codes under circuit-level noise, obtaining thresholds comparable to those of rotated surface codes.

Large language models (LLMs) are increasingly being used to support computer science research. To contribute reliably, these systems must be able to work with information distributed across research papers and across different aspects of the research process. However, large-scale resources that systematically organize such information into structured supervision for both model evaluation and post-training remain limited. To address this gap, we propose a systematic paper-to-QA construction pipeline that organizes information from AI research papers into structured QA pairs supported by source evidence. Using this pipeline, we construct CS-54k, a 54k-pair corpus derived from 14,474 openly licensed AI research papers. The corpus covers eight recurring paper components, including research domain, prior work, existing challenges, motivation, findings and assumptions, methods, experimental settings, and experimental results. We further derive two complementary subsets: CS-4k, a benchmark for evaluating LLMs on paper-derived research QA, and CS-50k, a companion corpus for post-training. Extensive evaluations on CS-4k reveal substantial performance differences across existing LLMs and paper components, with experimental settings and results remaining particularly challenging. Moreover, post-training on CS-50k substantially improves Qwen2.5-7B-Instruct on CS-4k, demonstrating the practical utility of structured, paper-derived supervision. Together, CS-4k and CS-50k provide a unified and scalable resource for benchmarking and improving LLMs on research paper question answering.

The increasing penetration of inverter-based resources and growing cyber-physical interdependencies expose electrical power systems to disturbances that challenge conventional resilience assessments. Existing approaches often evaluate resilience dimensions independently, overlooking the non-additive effects of coordinated disruptions and the influence of exogenous stressors. To address this gap, this paper proposes a Grid Multidimensional Resilience Index (GRID-MRI) that integrates endogenous disruption severity, coordination effects, and climatic and economic-regulatory conditions. The framework quantifies physical, cyber, and operational degradation, identifies interaction effects through counterfactual (i.e., what-if), simulations of individual and coordinated attacks, and incorporates exogenous stressors as scenario-dependent scaling factors. Its application to a modified IEEE 39-bus system under three escalating cyber-physical attack scenarios, informed by the December 2025 Polish cyberattack incident, reveals that coordinated attacks induce system collapse despite stable individual and additive responses. The corresponding resilience losses are 3.72 and 4.09 times their additive estimates, while GRID-MRI increases from 0.062 to 1.188 and reaches 6.081 under additional exogenous stress. These findings demonstrate the limitations of additive assessments and establish a quantitative basis for distinguishing the dominant mechanisms of multidimensional resilience degradation.

Existing natural-language-to-SQL benchmarks primarily evaluate queries over bounded relational data, leaving streaming semantics underexplored. We introduce FlinkEval, an execution-grounded synthetic benchmark comprising 3,903 question--SQL pairs across 45 industry domains and 45 semantic labels. We create diverse schemas and domain-plausible, temporally consistent data through iterative generation and validation. SQL patterns collected from the human-written NL2SQL pairs and vendor documentation drive quota-balanced generation, and execution with a repair loop on Confluent Cloud Apache Flink produces reference queries covering windowing, temporal joins, pattern matching, nested types, and data manipulation. The corpus includes paired schema-explicit and business-language requests, with annotations for analysis by query pattern, complexity, and execution mode. A seven-level Question Specification Ladder examines semantics-preserving reformulation and underspecified requests augmented with oracle-derived clues. Baseline evaluations combine execution outcomes with adjudicated LLM judgments, revealing challenges in pattern composition and request specification.

Provably secure linguistic steganography assumes that the receiver recovers the sender's tokens, but a text channel delivers only text, and ambiguous tokenizations break extraction. Existing fixes change the distribution, synchronize whole pools of tokens at a capacity that collapses with large vocabularies, or track several token histories. We prove that if extraction is exact and error-free across full-support language models, a single-history receiver is confined to a prefix-free candidate set at every step -- even if it reads the entire future text and the sender constrains later steps -- and must therefore give up the stop entropy, the entropy of deciding whether each prefix token ends there. Fixing these stop decisions on the prefix forest with shared randomness yields coin projection: a closed-form scheme that attains the bound, works with any distribution coder in one model call per token, preserves ordinary sampling, and refines SyncPool for every key. Across English, Chinese and Japanese, ambiguous text is rare (at most 11 in 1,000 tokens), yet the stop entropy at word and morpheme prefixes takes 7.2-9.0% of the entropy at every step; SyncPool's further loss comes from space and single-byte UTF-8 hubs that merge subtrees while carrying almost no stop uncertainty. Coin projection keeps 91-93% of the string entropy for the coder at every vocabulary size; with the full vocabulary, it turns 2.16-6.85 times as much entropy into message bits as SyncPool and recovers every completed message.

In the Online Transportation problem, introduced by Kalyanasundaram and Pruhs (1995), requests arrive sequentially in a metric space and must be assigned immediately and irrevocably to one of $m$ facilities with available capacity. The objective is to minimize the total distance between requests and their assigned facilities. The problem generalizes the Online Metric Matching problem, in which each facility has unit capacity, and inherits a $2m-1$ lower bound on the competitive ratio for all deterministic algorithms. For Online Metric Matching, Kalyanasundaram and Pruhs (1993) and Khuller, Mitchell, and Vazirani (1994) established an optimal deterministic competitive ratio of $2m-1$, attained by the Retrospective (Permutation) algorithm. Kalyanasundaram and Pruhs (1998) conjectured that the same ratio is achievable for online transportation. Recently, Seth, Das, and Raghvendra (2026) obtained a competitive ratio of approximately $6.66m-2.89$ using their Robustified Greedy algorithm. We give an algorithm with competitive ratio $ρ(m-1)+1$, where $ρ\approx5.64$. Our algorithm builds on minimum-spanning-tree search techniques developed by Harada and Itoh (2025) and Arndt, Moseley, Pruhs, and Uetz (2026), using priority-based virtual depth-first searches over the known facilities. However our algorithm uses a phase to shift the geometric search scales, to decide when to skip certain searches and optimize the thresholds that define what edges to include in a given search.

To meet the throughput demands of blockchain systems, protocols for State Machine Replication increasingly have many processors disseminate transaction blocks in parallel. Certified designs require availability certificates before blocks enter the ordering process, which is robust but adds latency. BlueBottle, the leading uncertified design under $n\geq 5f+1$ ($n$ processors, at most $f$ Byzantine), reduces latency in favourable conditions but can put fetching missing data on the critical path. Raptr, the state of the art under $n\geq 3f+1$, avoids fetching on the voting path, but transactions wait for a leader's proposal and two voting rounds, and the leader controls the proposed sequence. We prove that strengthening the resilience assumption alone cannot remove Raptr's second voting round. We present Multimmit, a protocol for $n\geq 5f+1$ that builds on Raptr's approach and finalises leader blocks after a single round of voting, improving latency by at least one message delay. Extension votes let voters endorse fresh transaction blocks omitted from the leader's proposal: once a leader block finalises, every fresh, well-circulated block of an honest chain enters the total ordering by the next finalised leader block, even if that block's leader is Byzantine. Producer chains are interleaved by a prescribed rule, not in a leader-chosen sequence. Payments are final within a single round trip, once the producer assembles $n-f$ acknowledgments into a transferable receipt. With fifty validators across thirteen regions, Multimmit's median latency to placement in the total ordering moves from 383 to 411 ms as the offered rate rises from 100k to 1M transactions per second. At 100k, it is essentially unchanged under 0.1% message loss and rises by a factor of 1.4 with nine validators crashed; these conditions slow BlueBottle and Raptr, respectively, by more than an order of magnitude.