Papers

As IT systems grow in complexity, automated agent-driven incident response becomes increasingly critical. A key capability for these agents is summarizing large log streams into human-readable reports. Further progress requires meaningfully evaluating this function, yet existing benchmarks leave end-to-end summarization from large log streams underexplored due to the prohibitive cost and complexity of collecting real-world data. We introduce SLS-Bench, a benchmark for incident log summarization featuring 90 tasks grounded in real-world incident reports, each with a synthetic log stream and a reference summary. To construct SLS-Bench, we developed a data generation pipeline that converts incident reports into structured incident models, which guide the generation of synthetic logs and reference summaries. This pipeline combines narrative grounding, explicit incident modeling, generator-verifier loops, and statistical comparisons with real-world log streams to produce realistic and controllable observability data. Our evaluation of 14 LLMs on SLS-Bench shows that while frontier models can produce useful summaries, they do so at substantial cost and latency, and open-weight models lag significantly. SLS-Bench provides a framework for systematic evaluation of log analysis agents and a scalable blueprint for future synthetic observability benchmarks.

Deploying robots in human-shared environments requires deciding not only how to navigate, but also when dedicated social signals are worth emitting. Classic observation-based inference can be misleading in such settings: confounding factors may bias the estimated relationship between a signal and the task outcome, leading to incorrect decisions. Causal inference provides a framework for disentangling these relationships by modelling the mechanisms linking observations, actions, and outcomes, and by estimating interventional effects after accounting for confounding paths. We study a human-robot interaction scenario in a corridor and use a causal model to estimate whether a social navigation signal causally improves transit success. We represent the interaction as a causal DAG involving robot action, pedestrian occupancy, available space, task outcome, and a controlled confounder (static obstacles). We compare the interventional estimate with the observational estimate obtained by conditioning only on the robot action, which does not explicitly model interventions. Our results show that accounting for confounding variables can reverse the conclusions drawn from naive observational analysis: the observational estimate wrongly suggests that the social signal is detrimental, whereas the causal estimate shows that it is beneficial. This discrepancy highlights how failing to account for confounding can lead to incorrect conclusions and demonstrates the importance of causal models for context-aware signalling decisions in social robot navigation.

With AI automating consequential tasks, human oversight is required for ethical, legal, and practical reasons. Yet these oversight tasks can be demanding and often make for poor work. Empirical evidence on overseers' lived work experiences remains underexplored, yet critical for designing “good” oversight. We provide longitudinal, in-the-wild insights through action research accompanying (N=8) tutors who oversaw an AI-supported learning assistant over three months. Our thematic analysis combines insights from twelve weekly meetings and two workshop iterations and was enriched by four repeated questionnaires. We identified stimulation, mastery, and relatedness as key challenges, stemming from the task's repetitiveness, unclear impact of interventions, and limited connections with students and peers. Informed by our initial findings, we deployed targeted UI modifications with mixed success. This paper documents the long-term evolution of the work characteristics of a real-world oversight task and discusses which oversight work characteristics can and cannot be supported by interface design.

Repetitive structures can give an incorrect pose candidate more inlier support than an accurate one, causing visual localization to select it. History-Guided Pose Reselection (HPR) uses coarse position forecasts from up to five past localization outputs to distinguish well-separated current-query candidates. The best-fitting forecast proposes an alternative; HPR accepts it only if every well-fitting forecast is much closer to it than to the localizer's pose. HPR needs no odometry and returns the selected candidate unchanged, using prediction for selection rather than pose refinement. With one development-set threshold, HPR improves recall within 5 meters and 10 degrees in all 12 main configurations of Hierarchical Localization (HLoc), MASt3R-based pose generation, and JIST-based retrieval, losing successes on fewer than 0.4 percent of queries. On NAVER Gangnam B1, HPR corrects 97 HLoc failures and loses six successes; gains persist within 0.25 meters and 2 degrees on all three HLoc test sets. Median HPR runtime on HLoc candidates is below one millisecond, excluding pose generation. Recovery weakens under less frequent localization.

Opinion dynamics models relate interpersonal influence to consensus or persistent disagreement. We study a discrete-time model combining weighted-average and weighted-median updates on a fixed directed network. Individuals update synchronously with possibly different adjustment coefficients. We characterize equilibria through local balance conditions and opinion level sets, and establish convergence to a Lyapunov-stable equilibrium using nonexpansiveness and Krasnosel'skii–Mann theory. We define type-dependent cohesive sets: each weighted-average member assigns all its weight to the set, whereas each weighted-median member assigns at least one half. Consensus is reached from every initial condition if and only if no two such sets are disjoint. Otherwise, an open set of initial conditions leads to persistent disagreement. Replacing weighted-median individuals by weighted-average individuals preserves consensus whenever it is already guaranteed for all initial conditions. Four numerical studies examine the number and locations of weighted-average individuals. The frequency of reaching a prescribed consensus tolerance is nondecreasing with their fraction in the sampled ensembles, but individual realizations need not follow this trend.

Web application development often proceeds through successive requests that extend an existing implementation. Yet web-generation benchmarks primarily evaluate development toward fixed specifications. We introduce EvoWeb, a novel benchmark comprising 320 cumulative requests over 32 web applications, grounded in real-world product features and structured by functional dependency graphs. Agents iteratively modify a persistent codebase, while a multimodal web agent tests each intermediate application in a live browser, assessing newly requested functionality, cross-request integration, and previously requested functionality. To characterize reliability throughout development, we integrate trajectory performance across user-tolerance thresholds rather than impose a fixed stopping criterion. Experiments across open- and closed-source frontier models show that EvoWeb reveals clear performance differences between models, with cross-request integration emerging as a key challenge. Further analysis shows that later requests with deeper or more numerous dependencies are solved less accurately, while runtime errors become more frequent over successive requests. EvoWeb provides a testbed for evaluating and improving coding agents in long-horizon, multi-request development.

Robot learning has undergone significant advancements in the domain of compliant control and robotic assembly. The development of novel methodologies has transcended the capacity to address discrete tasks, such as grasping or inserting, and has evolved to encompass the comprehensive construction of entire assemblies. Conventional benchmarks, which are predominantly focused on single tasks, are no longer adequate to address the complexity and interconnected nature of modern manufacturing processes. Consequently, the necessity for benchmarks to encompass a broader spectrum of assembly and subassembly steps is becoming apparent. In light of these developments, we propose LABIT, an industrial robotic assembly benchmark for long-horizon manipulation. This benchmark encompasses DIN-norm assembly tasks, including screwing, placing, inserting, and snapping, under both high and low tolerance conditions. The successive nature of these assembly steps implies that prior errors can influence subsequent manipulation. Consequently, this benchmark is particularly relevant for assessing the robustness of intelligent robotic systems. The proposed benchmarking protocols and novel performance metric, based on a taxonomy of assembly complexity in conjunction with the comprehensive list of parts and their corresponding CAD models, serve as essential tools for research and development in the field. To provide a baseline, we evaluate the benchmark using traditional force-position control and report reliability and score metrics per primitive, subtask, assembly category, and subassembly. Accompanying data is available at https://irp-hka.github.io/robotic-assembly-benchmark/.

We study environment scaling as a way to increase agent-benchmark difficulty while holding core task objectives fixed. Our framework, Recursive Environment Scaling (RES), transforms existing benchmark environments while preserving their objectives and completion requirements. A proposer changes information access, resource relations, and execution dependencies; execution-grounded checks validate the resulting candidate, and accepted changes are inherited across rounds. This produces related environments for paired evaluation and training-trajectory collection. Experiments on 294 initial tasks from three benchmarks show lower task success for all six evaluation models, with decreases of 57.14–77.14 percentage points on Claw-Eval-Live. With equal trajectory counts and identical optimization settings, fine-tuning on RES trajectories outperforms original-trajectory SFT in all nine model–benchmark comparisons by 7.00–14.61 percentage points. Ablations identify recursive inheritance and dependency composition as important contributors to increased difficulty. Together, these results support environment scaling for both agent evaluation and training.

Guided sampling with sparse observations and PDE residuals has recently been combined with neural operator priors to solve partial differential equations in function space, their natural setting. We pair such a prior with pseudo-bootstrap sequential Monte Carlo (pBS SMC), a guided sampling framework that accommodates proposals without tractable densities, and introduce a new proposal by adapting the DPM++ algorithm to the guidance setting. Across five benchmark PDE systems this proposal attains the best error–time trade-off at every step count tested, and wrapping it in pBS gives the lowest error on most inverse-problem settings, with the largest gains when few sampling steps are used.

The fusion of artificial intelligence (AI) and space technologies is extending spaceborne intelligence beyond onboard payloads toward a global infrastructure, here termed Space AI infrastructure (SpaceAI-Infra). Its promise rests on two distinctive advantages of space systems: abundant solar energy in orbit and the global reach of satellite constellations. Realizing this promise hinges on a fundamental systems question: How should satellite constellations be designed and orchestrated to deliver data-center-scale AI services competitive with their terrestrial counterparts? This article develops a roadmap-oriented answer. We first chart the evolution of SpaceAI-Infra through three progressive stages: onboard AI, Space Edge AI, and Space Cloud AI. By revisiting the space-constrained fundamentals for end-to-end token services, we derive enabling design principles and translate them into an integrated space edge–cloud architecture. The communication-centric space edge comprises free-flying LEO constellations that provide user-proximate computing and form the access and backhaul fabric for cloud AI services. Complementing the edge, the computing-centric space cloud comprises formation-flying satellite constellations organized as space data centers in dawn–dusk sun-synchronous orbits. A case study demonstrates that the proposed architecture can support data-center-scale foundation-model inference for millions of concurrent users worldwide.

We study the following simultaneous embedding problem: Given k graphs on the same vertex set with the same pairwise intersection, do they admit planar drawings that coincide on their shared graph? We establish fixed-parameter tractability with respect to the vertex integrity of the union graph (that is, the union of the input graphs) plus k, resolving an open question and generalizing the known algorithm parameterized by vertex cover number plus k. Our main technical contribution on the way towards this result is the extension of the recently introduced Ramsey pruning technique beyond ordering problems. Finally, we complement our tractability result with a lower bound that establishes NP-hardness when both k and the pathwidth of the union graph are bounded by constants, ruling out a generalization of our main result to path- or treewidth.

Multi-step retrosynthesis planning is often considered solved once a route to purchasable starting materials is found, but such a route should also be practical in real-world chemistry. Recent LLM-based planners formulate this process as sequential decision-making over an AND-OR graph, yet typically expand a single molecule per turn and are optimized for solvability alone, leaving alternative routes uncompared across the multiple objectives that determine practicality before the search budget is committed. We introduce RetroPlanner, an LLM-based planner that jointly compares multiple routes to steer the search toward higher synthesizability, as measured by reaction plausibility, round-trip consistency, and starting-material price. RetroPlanner is trained to reason over chemical evidence: structural progress from the target molecule, bond changes and reaction classes, where candidate routes diverge from one another, and how they rank across multiple objectives. Experiments on multi-step retrosynthesis benchmarks demonstrate that RetroPlanner recovers more alternative routes than prior LLM-based planners while improving synthetic route quality and maintaining a high success rate.

A central goal of quantum machine learning is to identify applications in which small quantum circuits can improve task performance while remaining executable on real hardware. We investigate deepfake (synthetic) speech detection, a largely unexplored application in this area, where reliable confidence scores are as important as discrimination performance because deployment decisions are made at fixed thresholds and overconfident errors can be costly. We introduce QBFF, which routes a compressed speech representation through a small 8-qubit variational circuit (VQC) used as a drop-in bottleneck, with a matched classical bottleneck as the control. The quantum circuit yields better-calibrated confidence in most matched comparisons, including out-of-distribution evaluation, though the downstream deployment-cost benefit is model-dependent. We then run the same circuit on a real IBM quantum processor. To our knowledge, this is the first anti-spoofing pipeline to run on a QPU; hardware execution demonstrates feasibility, with lower subset EER and $C_{llr}$ but higher mean actDCF than the matched classical reference. Together, these results show that hardware-executable quantum bottlenecks can improve score reliability, with gains that vary by evaluation setting.

Allreduce is a fundamental communication primitive in large-scale distributed training, and torus interconnects remain widely deployed in high-performance systems. We introduce SP-allreduce, a new allreduce algorithm for $d$-dimensional tori that exploits the symmetry of the topology and routes only along shortest paths. We call an allreduce algorithm fully bandwidth-optimal if it incurs no bandwidth overhead, no multihop-induced traffic inflation, and achieves perfect per-round load balance. SP-allreduce is based on a family of carefully symmetrized shortest-path trees that guarantees full bandwidth optimality in the $\alpha$–$\beta$ model. We prove that it uses the minimum possible number of rounds among all fully bandwidth-optimal allreduce algorithms on $d$-dimensional tori, and therefore the minimum running time in this class. In this sense, SP-allreduce refines Bucket allreduce: it preserves Bucket's key strengths while reducing the number of rounds by a factor of two. We further extend our approach to broader symmetric topologies and to communication trees of prescribed depths. Experimentally, the resulting algorithms improve over the best compared state-of-the-art algorithms for tori by up to $1.19\times$ on square tori and $1.65\times$ on rectangular tori across the tested configurations. On HyperX and spine-leaf networks, we demonstrate gains of up to $14.82\times$ and $1.72\times$, respectively, over the best evaluated baselines.

Road-network trajectory synthesis under local differential privacy (LDP) enables mobility data analysis without uploading raw trajectories to an untrusted server. Existing locally private trajectory synthesis methods mainly model whole trajectories, point sequences, or grid-level mobility patterns. However, they cannot handle non-uniform privacy risks caused by local road-network and trajectory contexts, such as endpoints, long-stay-associated regions, and low-out-degree branches. They also do not directly protect conditional next-hop choices on road networks, where a user's selected successor may reveal sensitive local movement intentions. To address these challenges, we propose TrajRACE, a locally private synthesis framework for road-network trajectories. TrajRACE uses start, segment-count, and transition reports to represent trajectories. For each transition, it keeps the current road segment as a public aggregation context and perturbs the true successor among legal successors on the client side. It then assigns risk-aware protection strengths to transitions according to local road-network and trajectory contexts. Finally, the server recovers population-level start, segment-count, and conditional transition statistics from heterogeneous privatized reports, and generates synthetic trajectories that follow the public road-network topology. Experiments on three road-network trajectory datasets demonstrate the effectiveness of TrajRACE in improving synthetic-data utility across distributional, query-based, and pattern-level metrics, while providing stronger protection for high-risk local transitions.

Elite athletics throwing provides a structured biomechanical action sequence together with an objective continuous outcome, namely throw distance, making it a suitable setting for performance-oriented sports analytics. In this work, we investigate whether explicit phase alignment of pose dynamics improves throw-distance prediction under realistic broadcast conditions. To support this study, we curate Throwing4, a dataset of 500 elite throwing clips spanning javelin, discus, hammer, and shot put, with corresponding two-dimensional pose sequences, frame-level phase annotations, and official distance labels. We formulate throw-distance prediction from broadcast-derived pose as a regression problem and compare full-sequence modeling against a phase-conditioned approach that partitions each attempt into four technique stages and fuses phase-specific motion representations. Experiments across multiple graph-based pose encoders show a consistent reduction in prediction error when phase structure is incorporated, indicating that phase-aligned modeling provides a more stable representation than treating the entire motion as an unstructured sequence. These findings suggest that explicit biomechanical phase information is a useful inductive prior for pose-based performance modeling in elite athletics throwing. Beyond score prediction, the dataset and protocol provide a basis for future work on fine-grained technique analysis, interpretable sports analytics, and robust modeling from broadcast video.

Behavior-conditioned animal vocalization generation aims to synthesize vocalizations from natural-language descriptions of animal behaviors, social interactions, and contexts. This task remains largely unexplored, with limited behavioral text supervision and substantial variation in vocalization duration. We construct the Behavior-Vocalization Dataset (BeVo) by integrating public bioacoustic data from six species with natural-language behavioral conditions derived from species-specific ethological and bioacoustic literature, totaling 31,191 events. We then adapt a rectified-flow-based generator with Event-Aware variable-length modeling, which encodes vocalizations at their actual durations and uses masks to prevent attention to padded latent positions and exclude these positions from the training loss. We evaluate vocalization discriminability, acoustic distribution fidelity, and language-aligned representation fidelity. Experiments show that task-specific adaptation improves all evaluated metrics compared with its zero-shot baseline, while Event-Aware modeling further increases vocalization-anchor macro accuracy from 59.86% to 61.99% and reduces Frechet Audio Distance from 1.40 to 1.22. Conditioning analyses further show that behavioral conditions improve target-anchor discriminability, while swapping behavior shifts generations toward the corresponding vocalization anchors. Demos are available on our project page.

Using the formalism of presheaf automata, we introduce new classes of automata for modelling concurrency that extend higher-dimensional automata and symmetric higher-dimensional automata. We study their languages and characterise these languages in terms of posetal cospans.

We propose Pathfinder-BP, a recursive decoder for topological quantum codes that combines belief propagation with controlled bit-flipping. An initial BP pass identifies and permanently fixes high-confidence errors. Upon convergence failure, the decoder recursively explores alternative error patterns by tentatively flipping low-confidence variables. On surface codes, Pathfinder-BP achieves a noise threshold and logical error rate comparable to minimum-weight perfect matching (MWPM). A crucial advantage of Pathfinder-BP is its highly parallelizable structure, which yields a worst-case decoding latency quasi-linear in the code distance, making it a suitable candidate for low-latency hardware implementation.

Nominal chip specifications do not capture the compute deliverable by heterogeneous distributed networks under communication, runtime and verification constraints. We formalize D-PIN Standardized Compute Units (D-PIN SCU) as a dynamic metering system for deliverable compute. The system combines five-dimensional resource normalization with topology-aware and runtime-corrected metering. A cluster-scale Roofline model yields a closed-form communication ridge, while runtime factors describe thermal, fault and load effects. We show that runtime deviations cannot be identified from nominal specifications alone. Uniform re-anchoring preserves relative metering results, and a simplex-constrained calibration algorithm estimates resource weights. The framework couples Proof-of-Physical-Work verification to a no-arbitrage condition for over-claiming and electricity-aware pricing. An end-to-end recomputation on the Zhangjiang Science Gate production node covers ten heterogeneous 8-GPU servers totaling 59.5 D-PIN SCU. Billed throughput of approximately 33,000 D-PIN SCU-hours is consistent with the physical upper bound at 78% utilization over 720 hours. Derived metrics are 3.02 kW per D-PIN SCU at the supply-and-distribution design boundary and 0.26 D-PIN SCU-hours per delivered kWh. Empirically calibrated chip values differ materially from datasheet projections; the published calibration is not independently reproducible from vendor specifications. Validation is limited to one production node, and the originality claim concerns the integrated system rather than the SCU terminology or individual components. The framework provides a basis for scheduling, verification and pricing of deliverable compute in distributed heterogeneous networks.

Smart meters are pivotal to the functioning of smart grids, as they measure and transmit energy consumption data that includes sensitive information about household energy usage patterns. While this data offers significant financial benefits, it can also reveal detailed insights into residents' daily routines, making the security of smart meters a critical concern. This paper introduces a novel remote attestation method for smart meters that ensures the integrity and privacy of data exchanged between smart meters and central servers while detecting unauthorized tampering. Our approach structures the communication data into a Merkle Tree-a hash data structure that facilitates efficient data integrity and authenticity verification. All operations involved in constructing the Merkle root and hashing critical data are executed in the secure environment of a Trusted Execution Environment (TEE) using OP-TEE, thereby maintaining the integrity and security of the smart meter's data. Comparing our experimental results with the related works highlights the superiority of our method in terms of resource efficiency, security robustness, and the feasibility of implementation on smart meters and resource-constrained devices, thereby enhancing the overall security and reliability of smart grid systems.

Although users may express different personalization needs across tasks, these task-specific behaviors can reflect shared underlying preference mechanisms. Existing LLM personalization methods often focus on surface-level interactions or undifferentiated histories, overlooking the shared mechanisms underlying these behaviors. We introduce PersonaFactors, a framework that distills reusable and interpretable personalization factors from heterogeneous user interactions. We construct a library of thirteen PersonaFactors by extracting evidence-grounded rationales and clustering recurring patterns across tasks, and develop a dynamic factor-guided generation framework that selects factors relevant to each user's history and current request. Across four personalization benchmarks and four backbone models, Dynamic PersonaFactor achieves the best score on every reported metric. Representation analyses further show that PersonaFactors and user history induce partially overlapping low-dimensional representations, and that their joint use yields complementary shifts. Taken together, these findings suggest that factor-based and history-based personalization can mutually reinforce each other.

Large language models are increasingly deployed in morally and legally sensitive contexts, yet little is known about how they internally represent distinctions between moral reasoning styles. We ask whether a model's differential response to rule-framed and consequence-framed moral questions reflects distinct internal structure or merely surface lexical sensitivity. We construct a 2$\times$2 factorial dataset crossing reasoning style (rule-based vs. consequence-based framing) with judgment polarity (permissive vs. prohibitive), drawn from three established moral reasoning benchmarks and evaluated on a cleanroom test set of novel phrasings that includes conflict pairs where the two framings predict opposite labels. Using linear probing and activation steering, we identify two orthogonal directions in the residual stream: Dir-A tracks reasoning style and Dir-B tracks judgment polarity, each near-blind to the other factor (cross-correlation $|r| < 0.04$). This double dissociation replicates across three instruction-tuned models from different developers at comparable proportional depth, and it is not knife-edged in layer: the directions are recoverable and causally effective throughout an extended band of the network. Dir-A can causally change judgments, and in conflicting scenarios, it selectively corrects errors in rule-based judgments without changing consequence-based judgments, suggesting that it does not simply make the model more likely to accept everything. Generalization beyond the training templates is partial: the direction transfers across benchmark sources, but its advantage over lexical baselines on novel phrasings is modest. We therefore frame the result as moral reasoning style sensitivity, a geometrically dissociable and causally effective structure, rather than internalized normative reasoning.

Language agents increasingly modify robot software, but whether they can repair it is unmeasured, because the domain supplies no oracle to inherit: a robot bug manifests as behavior, and robot repositories ship almost no tests that capture it. We present RoboBugBench, a benchmark of 33 real, execution-validated defects in a pinned ROS 2 navigation stack, in which a candidate patch is graded by compiling it and driving the robot through a physics simulation rather than by running a unit test. Three frontier models, each given 1 attempt per instance, repair 24.2%-36.4%, with no pairwise difference significant; difficulty tracks the oracle rather than the domain, at 66.7% on instances graded by process survival against 22.2% on those graded by what the robot did (instance-level p = 0.0027). Because the oracle is constructed rather than inherited we audit it adversarially, and report the decomposition that audit yields: counterfeit patches which suppress a failure without repairing the defect are accepted by the shipped rule 11/11, while a distributional rule over the same data rejects 4 at no cost in false rejects. The residual is an observability bound (what the observation channel discards, no predicate over it can recover), so discrimination must be bought with instrumentation rather than stricter criteria, and our repair rates are bounds: inflated by what the oracle cannot distinguish, deflated by unanimity's rejection of correct fixes under run-to-run noise. We release the corpus, criterion batteries, counterfeit suite and cached telemetry.

Looped Language Models (LoopLMs) improve parameter efficiency by recursively reusing shared Transformer modules, yet their repeated computation over evolving recurrent states poses new challenges for low-bit deployment. Existing post-training quantization (PTQ) methods, largely designed for feed-forward Transformers, fail to account for cross-loop activation dynamics, heterogeneous loop sensitivity, and recursive error propagation, leading to substantial performance degradation at low precision. To address these challenges, we propose LoopQuant, a loop-aware PTQ framework for accurate and efficient low-bit quantization of LoopLMs. LoopQuant is built on three core components: 1) Cross-Loop Distribution Canonicalization (CLDC) learns a shared quantization-friendly representation across recurrent states while preserving parameter sharing; 2) Interaction-Aware Loop-wise Precision Allocation (ILPA) allocates activation precision under a fixed bit budget by modeling task-level loop sensitivity and cross-loop interactions; and 3) Loop-Conditioned Task Calibration (LCTC) further aligns quantized inference with the low-bit recursive trajectory through lightweight task-space calibration. Extensive experiments across LoopLM architectures demonstrate that LoopQuant consistently outperforms existing low-bit PTQ methods. On Ouro-2.6B, LoopQuant under W4A4 surpasses all W4A8 baselines in performance, while reducing model memory by 65.3% with only 3.4% additional storage overhead over naive INT4 and achieving a 1.7 times decoding speedup over BF16. These results demonstrate a strong balance among accuracy, compression, and inference efficiency, providing a practical and deployment-friendly solution for low-bit LoopLM inference. The code is available at [https://github.com/Kishon-zzx/LoopQuant].

Video world models must generate physically plausible futures, not just realistic frames. Existing approaches typically use semantic, physical, or predictive information to guide the visual generation stream, leaving a fundamental question: where inside the generator should predictions of scene evolution live? Standard video tokens jointly encode appearance and dynamics, but provide no distinct state for learning and reading predictive dynamics. We answer this question with the Physical Registry: temporally grounded registry slots inside a video DiT that are initialized from an observation-conditioned forecast, co-evolve with video tokens, and receive future-representation supervision, making predictive dynamics an explicit, readable state of the generator. In a controlled comparison of predictive-state interfaces on Physics-IQ Verified, Physical Registry improves vanilla fine-tuning by 9.9/8.7 points on I2V/V2V, the largest gain among the interfaces we compare. Across all Cosmos and WAN backbones, Physical Registry improves Physics-IQ Verified scores. With only 0.45% additional parameters, it enables Cosmos 3 Nano 16B to surpass the published I2V score of Cosmos 3 Super 64B (43.56 vs. 42.7). On WorldModelBench, it also improves physical adherence and the total score for all four evaluated backbones. Because the registry becomes candidate-specific while remaining readable, the Physical Registry Selector (PRS) ranks emerging futures by how far their intermediate registry readouts depart from the shared forecast before full denoising is complete. Across four backbones, this selection improves I2V scores by 5.6-15.9% without completed-video external scoring. Together, these results support a broader principle: dynamics prediction should live as an explicit model state that co-evolves with the video representation. Project page: https://grainsack.github.io/physical-registry/

Speech-driven visual grounding in remote sensing (RSVG) removes the typing bottleneck of text-based interaction and is intended to preserve prosodic intent, enabling hands-free target acquisition in time-critical missions. The direct-audio paradigm, mapping raw speech straight to a bounding box and bypassing automatic speech recognition (ASR), was recently adopted by RSTalker, but what such grounders actually learn about language remains unmeasured. SemEcho grafts a frozen Whisper speech pathway onto a VLTVG grounding host so that speech and text share one decoder, making the paired text, free by construction since speech corpora in this task are synthesized from DIOR-RSVG expressions, a training-time teacher and a fair, same-host text upper bound. On this instrument we dissect three semantic injection channels (feature-mimicry distillation, word-level CTC supervision, and two-stage ASR pretraining) with linear semantic probes, and release Audio-DIOR-R (80,973 training clips; 10 timbres x K=3; Qwen3-TTS synthesis with online sox noise lambda in [0, 0.4]) and Audio-DIOR-Comp, a re-split holding out 17 attribute-category combinations under a 2x2 acoustic-semantic protocol. Our anatomy yields four findings. (i) Whether direct alignment beats a domain-trained ASR cascade is decided by recognizer quality: against a weak in-domain cascade (CTC 0.53) direct wins by 9.0 Acc@0.5 points with transcription errors concentrated on spatial-relation words, but a strong in-domain cascade (CTC 0.029) reaches the text upper bound (80.8 vs. 81.0) and beats even the strongest direct model by 8.8 points; the anti-ASR premise holds only against weak recognizers.

A decoding method is proposed for polar codes under channels with strong phase noise. In contrast to traditional SC decoding, where recursions are performed on scalar soft information such as probabilities, the proposed decoding recurses over probability density functions defined on the phase-noise variable. To make the density-level recursion computationally tractable, we represent each log-density by its spectrum, and derive approximate formulas for the spectrum recursion, therefore, the proposed method is termed log-density spectrum recursion (LDSR). Numerical results are given under strong phase noise and AWGN, showing that a prominent superiority of error performance is achieved by LDSR in comparison with traditional SC and Kalman-based methods.

Multilingual translation requires a model to preserve meaning while producing text in a requested language. We present HexaBee-4B, a translator adapted from Gemma 3 4B IT through three stages: Wikipedia continued pretraining, joint multilingual answering and answer-text translation, and bidirectional translation specialization. The intermediate stage connects two tasks through shared multilingual answers: answering a question in the requested language and translating an answer into that language. It comprises 19.46 million answering records, including 5.76 million cross-language question-answer pairs, and 10.70 million answer-text translations. The final stage uses 658,274 forward and reverse synthetic translation examples. On 36,572 synthetic test examples spanning all 30 directions among Arabic, Chinese, English, Hebrew, Persian, and Spanish, HexaBee-4B achieves direction-macro chrF++ of 73.75, compared with 57.11 for Gemma 3 4B IT and 66.83 for Gemma 3 12B IT. BLEU, TER, COMET, and a matched model-judged subset support the same overall ranking. These results demonstrate improved translation across the six-language set, with chrF++ gains in every evaluated direction over both baselines.

We revisit the classical computational geometry problem of computing the exact (discrete or continuous) Fréchet distance between two given polygonal curves with $n$ vertices in a constant dimension $d$. 1. We present an $O((n^2/\log^2n)\log^2\log n)$-time deterministic algorithm for the discrete Fréchet distance problem, improving the previous $O((n^2/\log n)\log\log n)$ result by Agarwal, Ben Avraham, Kaplan, and Sharir [SODA'13]. 2. We present an $O((n^2/\log^2n)\log^2\log n)$-time deterministic algorithm for the decision version of the continuous Fréchet distance problem, improving the previous $O((n^2/\log n)\log^2\log n)$ result by Buchin, Buchin, Meulemans, and Mulzer [SODA'14] and the recent $O(n^2/\log^{1+\mu}n)$ randomized result by Cheng and Huang [SODA'25] for a small unspecified constant $\mu>0$ (which converges to 0 as $d$ increases). 3. Finally, we present an $O((n^2/\log^2 n)\log^{O(1)}\log n)$-time Las Vegas randomized algorithm for the continuous Fréchet distance problem, improving the previous $O(n^2\log^2\log n)$ randomized result by Buchin, Buchin, Meulemans, and Mulzer [SODA'14] and the recent $O(n^2/\log^{\mu}n)$-time randomized result claimed by Cheng and Huang [SODA'25]. Since these problems have near-quadratic OV-based conditional lower bounds and shaving of two log factors is the best that is known for similar, simpler problems such as longest common subsequence and edit distance, our results reach the natural limit of the current state of the art (up to $\log\log$ factors).