Papers

Embodied Question Answering (EQA) requires an agent to explore a previously unseen environment, gather relevant information, and answer questions about the scene. Recent approaches leverage Vision-Language Models (VLMs) together with semantic maps or scene graphs to guide exploration. However, exploration is typically driven only by local observations, while structural priors about the environment remain largely unused. We propose HFLEX-EQA, a hierarchical EQA framework that combines online scene graph construction, VLM- based planning, semantic frontier exploration, and floorplan priors. The system incrementally builds a hierarchical scene graph and an open-vocabulary occupancy map from RGB-D observations, enabling a VLM to jointly reason over the scene graph, task-relevant visual observations, exploration history, and an estimated topological floorplan. Furthermore, we introduce a room-discovery strategy that leverages the floorplan and open-vocabulary frontier semantics to guide exploration toward semantically relevant yet currently unobserved room types. We evaluate HFLEX-EQA on the OpenEQA and ExploreEQA benchmarks and demonstrate deployment on a quadruped robot in real indoor environments. Our results demonstrate the benefit of combining VLM-based hierarchical planning with structural floorplan priors for the EQA task.

Tensor-based modulation (TBM) schemes are a promising approach for unsourced random access (URA), where user separation relies on decomposing the received signal tensor via the canonical polyadic decomposition (CPD), typically computed with alternating least squares (ALS). Standard ALS, however, treats the factor matrices as unstructured and fails to exploit the discrete structure of the tensor sub-constellations. We propose DVB-ALS, a discrete variational Bayesian CPD framework with specific priors, tailored to a tensor structure with the corresponding encoding strategy, combined with iterative computation of an approximate posterior distribution. A discrete Gaussian mixture prior on one Grassmannian factor softly aligns the estimates toward the constellation points. The remaining factors are jointly modeled with a structured Gaussian prior whose posterior mean is constrained to the Khatri-Rao product manifold and posterior variance upper-bounded to prevent norm divergence during inference. The resulting closed-form coordinate ascent algorithm jointly estimates all latent factors and their uncertainties. We integrate DVB-ALS into DVB-TBM to design a complete URA receiver with single-user demapping, polar decoding with cyclic redundancy check (CRC) verification, and successive interference cancellation (SIC). Simulation results show significant gains over standard ALS-based decomposition and robust detection performance in URA settings, outperforming state-of-the-art schemes under high system loads.

The Terrapin attack against SSH channel integrity (USENIX Security 2024) used a novel attack vector: attacks on the channel state. Surprisingly, not all AEAD modes of SSH were equally affected by this attack, and it remained an open question if "unaffected" meant "secure". Existing formal models for secure channels are based on stateful encryption. However, these models do not define what the channel state is and how it is used as input to the different AEAD modes. In this paper, we propose a formal model for channel integrity under partially chosen state. Applied to the Terrapin attack, the chosen state is the SSH sequence number. It uses an abstract stateful encryption interface, for which we provide pseudocode descriptions for the eight most prominent AEAD modes used in SSH. By varying the SND oracle, we can model ciphertext-only (CO; the Terrapin attack), known-plaintext (KPA), and chosen-plaintext (CPA) attacks. This allows us to establish concrete bounds on the security of the AEAD modes. We find that all three Encrypt-then-MAC (EtM) modes and ChaCha20-Poly1305 in SSH are insecure in the CO model. AES-GCM is the only cipher secure in all three model variants. Going beyond Terrapin, we show that Encrypt-and-MAC (EaM) with a CBC cipher is secure, even in the KPA model. In particular, we describe a novel BEAST-like chosen-plaintext attack on the channel integrity of EaM-CBC, which separates the KPA and CPA models for this scheme.

Reinforcement learning has become a central component of large language model (LLM) post-training, yet token-level credit lacks a generally accepted mathematical definition, leaving its relationship to commonly used training signals unclear. We formulate three regularity conditions, namely Completeness, Prefix Consistency, and Neutrality, and prove that they uniquely determine token-level credit. This characterization provides a unified basis for explaining phenomena across existing algorithms and guides the development of an improved actor-critic training procedure. Through this lens, an ideal teacher in On-Policy Distillation (OPD) acts as an implicit critic, yielding an expected policy gradient proportional to that induced by token-level credit. Response-level REINFORCE Leave-One-Out (RLOO) signals match the expected policy-gradient contribution of token-level credit despite their coarser granularity. We further establish approximate credit sparsity under bounded outcome rewards and show how intermediate critic errors in Generalized Advantage Estimation (GAE) can become comparable to the underlying credit. These motivate Policy Aligned Critic Training (PACT), which adopts an Actor-then-Critic update order to apply importance sampling correction to critic training and better align the critic with the updated policy. In agentic mathematical reasoning, PACT achieves 72.87% average accuracy across four benchmarks, outperforming GRPO and PPO by 8.80 and 13.16 percentage points, respectively. On SWE-bench Verified, PACT achieves a pass rate of 67.4%, outperforming PPO, GRPO, and SAO by 2.4, 2.0, and 3.8 percentage points, respectively.

The EU Digital Services Act (DSA) requires Very Large Online Platforms and Search Engines (VLOPs/VLOSEs) to assess and mitigate systemic risks, including actual or foreseeable negative effects on fundamental rights, and subjects these assessments to independent audit. Yet how such rights-relevant risks should be identified, assessed, evidenced, and documented remains methodologically under-specified. We address this gap by introducing a decision-tree-based toolkit that operationalizes Article 34(1)(b) DSA and related audit requirements into a structured assessment process. The toolkit decomposes platforms as systems-of-systems and guides assessors through DSA applicability, context definition, subsystem and stakeholder identification, risk identification, risk assessment, reporting, and visualization. A central design contribution is its treatment of vulnerability as situational and relational: rather than assigning vulnerability to fixed groups, the toolkit asks how particular platform subsystems and dependencies may place stakeholders in vulnerable positions. It further makes fundamental rights-relevant judgments traceable by recording the affected right, interference and justification analysis within the structure of a proportionality test, while also covering a perception and impact dimension including severity, scope, duration, reversibility, as well as practical consequences. We developed the toolkit iteratively through a design-science process combining literature-derived design requirements, three interdisciplinary expert workshops (N=4; N=11; N=20) testing the framework design with two DSA user persona scenarios. We contribute an extensible methodological foundation for making fundamental-rights risk assessments under the DSA more streamlined, systematic, transparent, and auditable.

Concept drift, driven by the rapid evolution of Android malware, severely degrades the performance of machine learning detectors. Current adaptation strategies are often reactive, responding only after performance has dropped and imposing a significant manual annotation burden, or they are proactive but rely on unstable adversarial training and incomplete, single-level graph representations. To overcome these limitations, we propose HYDRA (Hybrid Drift Adaptation), a proactive adaptation framework that learns drift-invariant representations from hierarchically structured data. HYDRA first models applications using a hybrid graph structure, combining fine-grained Control Flow Graphs (CFGs) and coarse-grained Function Call Graphs (FCGs) to capture comprehensive behavioral patterns. It then introduces a novel cross-domain contrastive learning objective that aligns historical (source) and new (target) data distributions. By generating pseudo-labels for unlabeled target samples, our method pulls representations of semantically similar applications together, regardless of their domain, within a single, stable optimization process. This approach unifies feature learning and domain alignment, eliminating the need for complex adversarial objectives. Extensive experiments on large-scale, time-ordered malware datasets demonstrate that HYDRA achieves substantially lower False Negative and False Positive Rates than state-of-the-art baselines while requiring up to 87.5% fewer labeled samples. Our work thus offers a robust and efficient solution to combat concept drift in security applications.

The scarcity of non-English language data in specialized domains significantly limits the development of effective Natural Language Processing (NLP) tools. We present TransBERT, a novel framework for pre-training language models using exclusively synthetically translated text, and introduce TransCorpus, a scalable translation toolkit. Focusing on the life sciences domain in French, our approach demonstrates that state-of-the-art performance on various downstream tasks can be achieved solely by leveraging synthetically translated data. We release the TransCorpus toolkit, the TransCorpus-bio-fr corpus (36.4GB of French life sciences text), TransBERT-bio-fr, its associated pre-trained language model and reproducible code for both pre-training and fine-tuning. Our results highlight the viability of synthetic translation in a high-resource translation direction for building high-quality NLP resources in low-resource language/domain pairs.

Political discourse is widely perceived to be growing more hostile, yet robust evidence remains scarce. This study examines blame attribution in the Danish Parliament from 1997 to 2026, combining a purpose-built classifier, BlameBERT (F1: 0.80), with multilevel statistical modeling. The classifier is constructed using an annotation-efficient pipeline for blame attribution in low-to-mid resource languages. The results reveal a banana-shaped trajectory, with blame declining until around 2016 before entering a significant and sustained increase in recent years (2019-2026). Government status consistently influenced blame attribution - an effect we term political contrasting - with opposition parties blaming substantially more than governing parties. This effect was moderated by ideology: The blame-dampening effect of governing was less pronounced among right-wing parties, and ideological extremity amplified blame more strongly on the right. In recent years, the interaction between political wing and ideological extremity intensified, suggesting an ideological hardening of the blame rhetoric concentrated on the right of the political spectrum. Taken together, these patterns suggest that the perceived rise in harsh political language reflects not merely a general rhetorical drift, but an ideologically asymmetric hardening of political discourse. A sensitivity analysis showed that the conclusions were robust to varying classification thresholds.

SMT solvers make automated verification convenient. At the same time, solvers suffer from instability, whereby seemingly inconsequential changes to the input may cause a previously quickly produced proof to fail or time out. This paper addresses a common cause of outcome instability (i.e., unsat/unknown fluctuations) in the context of program verification. We demonstrate that the generation (i.e., depth) accounting used to make quantifier instantiation practical and implemented in multiple state-of-the-art SMT solvers is non-confluent (i.e., prone to divergence), and that this leads to instability. We then address this deficiency and develop a new accounting method that is stable. The new method is justified using a sequence of refinements from an abstract, confluent solver model all the way to our implementation in Z3. Our empirical evaluation demonstrates that our implementation reduces outcome instability by 94% in the unstable core of the Mariposa benchmark without leading to performance regressions.

Residual Vector Quantization turns continuous representations into discrete, multi-level token sequences. Yet most methods operate in Euclidean space, despite the coarse-to-fine structure of the resulting codes and the latent hierarchies present in many data domains. Hyperbolic geometry offers a natural alternative for hierarchical representations, but naive hyperbolic extensions introduce geometric inconsistencies: non-associative hyperbolic addition prevents consistent residual aggregation, while standard straight-through gradient estimation ignores the geometry of the latent space. We propose a geometry-aware hyperbolic residual quantization that addresses these issues in both the forward and backward passes. In the forward pass, Hyperbolic Residual Aggregation restores the telescoping behavior of residual quantization on the Poincare ball. In the backward pass, a discounted Hyperbolic Straight-Through Estimator routes the reconstruction gradient through the quantizer as a single geometric block, avoiding unstable recursive gradient transport across residual stages. Evaluations on hierarchical prediction, recommendation, image tokenization, and neural audio coding tasks show that our method improves the stability and structural organization of hyperbolic residual codes over naive hyperbolic baselines. At the same time, we observe a clear structure-compression trade-off: Euclidean residual quantization remains preferable for pure compression, while geometry-aware hyperbolic quantization is most useful for hierarchically organized discrete latent spaces.

Data-calibrated stochastic MPC typically builds separate risk margins for many events along the horizon, such as times, facets, state/input components, or obstacles, and then combines them with a union bound. This approach is valid, but it does not match the key object used in the tube-MPC recursive-feasibility proof, which shifts a complete error tube. This paper develops collective tube MPC (CT-MPC), where the calibrated uncertainty object is the finite-horizon prediction-error trajectory. A reusable trajectory tube is calibrated offline, its cross-sections define deterministic Pontryagin tightenings online, and the certified violation event is that a fresh prediction-error trajectory leaves the tube. For linear systems with additive uncertainty and fixed ancillary feedback, we prove joint state-input safety over the prediction horizon, one-step recursive feasibility from an explicit shifted candidate, a finite-deployment risk bound, and a practical value-decrease inequality. The finite-sample certificate is distribution-free under split calibration and has beta-binomial form; its complexity is the certified number of residual trajectories that can define the tube, rather than the number of horizon-constraint blocks. We also give implementable shift-compatibility tests for polytopic tubes and a stable-compression fallback for irregular tube designers. Numerical experiments compare CT-MPC with Bonferroni tightening, a sample-envelope tube, and a joint-in-time conformal MPC baseline. The collective tube reduces deterministic tightening while preserving empirical safety and recursive feasibility.

Argument Mining (AM) transforms natural language into its underlying argument structures. This transformation is typically realized through a sequence of AM tasks that form an end-to-end AM pipeline. However, AM approaches often differ in how they conceptualize these tasks, making direct comparisons between them difficult and opaque. This calls for a more nuanced, task-level analysis of AM approaches to enable clearer comparison and assessment. This work presents a preliminary meta-study that systematically reviews several state-of-the-art end-to-end AM works and analyzes their pipelines through a triple-perspective framework—a linguistic, computational and domain perspective—to understand how the pipelines model arguments as structures, computes them, and integrates domain knowledge. We further propose a general design to the linguistic and computational perspectives, illustrating how key AM tasks are designed for modeling and computation of argument structures. Our proposed framework lays the groundwork for methodology-centered descriptions across AM approaches, facilitating deeper understanding and more systematic comparisons in future research.

Inductive definitions are an important form of knowledge in mathematics and computer science. Two common techniques to prove theorems about inductive definitions are the principle of mathematical induction and the principle of infinite descent. To formalize these principles, Brotherston and Simpson introduced the sequent calculus proof systems LKID, for mathematical induction, and LKID{\omega} and CLKID{\omega} , for infinite descent. LKID{\omega} is an infinitary system, in which proofs are infinite trees, and CLKID{\omega} a cyclic system, in which proofs are finite graphs. However, these calculi restrict to monotone definitions, while inductive definitions are generally non-monotone. The logic FO(ID) extends classical first-order logic with non-monotone inductive definitions. In earlier work, we provided a formalization of the principle of mathematical induction for non-monotone definitions by extending LKID to a sequent calculus SCFO(ID) for FO(ID). In this paper, we provide a formalization of the principle of infinite descent for non-monotone definitions by extending LKID{\omega} and CLKID{\omega} to sequent calculi SCFO(ID)-inf resp. SCFO(ID)-cyc for FO(ID). Furthermore, we extend several proof-theoretic results for LKID{\omega} and CLKID{\omega} to SCFO(ID)-inf and SCFO(ID)-cyc regarding soundness, completeness, cut-elimination and the relation with SCFO(ID).

The dimensional collapse of representations in self-supervised contrastive learning is an ever-present issue. One notable technique to prevent such a collapse of representations is using a multi-layered perceptron network called Projector. In several works, the projector has been found to heavily influence the quality of representations learned in a self-supervised contrastive pre-training task. However, the question still lingers. What role does the projector play? Assuming the projector mitigates dimensional collapse, what prevents the terminal layer of the base encoder from functioning as the projector in the absence of an explicit multi-layer perceptron (MLP) head? In this work, we intend to study what happens inside the projector by examining the rank dynamics of the same and the encoder through empirical study and analysis. Through mathematical analysis, we observe that the effect of rank reduction predominantly occurs in the last layer. Motivated by this insight, we propose a weight regularization strategy applied specifically to the last layer. We demonstrate that this targeted approach yields better performance than applying orthogonal weight regularization across the entire network (WeRank), both with and without a projector. Our method improves Top-1 accuracy by more than 1% on SimCLR on the ImageNet100 dataset and consistently outperforms baseline SimCLR variants on CIFAR datasets, supporting our interpretation of the projector's role.

Prefill and decode reward different approaches to quantization: low-precision arithmetic accelerates prompt processing, while compact weights reduce memory traffic during generation. We propose "disaggregated quantization" (DQ), which specializes computation formats, weights and storage placement to both of these phases. On Qwen 3 and Gemma 3, removing activation quantization specifically on decode improves accuracy on decode-heavy tasks without increasing inference cost. Training separate compute-native prefill weights accelerates prompt processing relative to weight-only inference while matching or exceeding its accuracy at 2-3-bit decode on both decode-heavy and prefill-heavy tasks. With released Qwen3.8-27B GGUF decoders, training an NVFP4 prefiller improves 1-bit accuracy by 32.5 points on MMLU-Pro and 35.3 on MMMU-Pro without modifying the decode checkpoint. To accommodate the additional checkpoint on a single device, offloaded disaggregated prefill (ODP) streams its weights from SSD, amortizing loading over prompt length. On the same 27B model, ODP delivers a 1.78x time-to-first-token speedup over the weight-only baseline at 8K prompt length in llama.cpp. We evaluate accuracy under disaggregated serving in vLLM and further validate shared-weight format disaggregation through post-training quantization on models up to 2.8T parameters.

We study preference elicitation under the Bradley-Terry-Luce (BTL) model where the true partworth vector is unknown and has to be estimated as a parameter with elicited preference information. The set of selected pairwise queries is non-uniform, deterministic, and arbitrary over a collection of alternatives, provided that it satisfies a joint identifiability condition. We focus on understanding when the canonical maximum likelihood estimator (MLE) is finite and admits sharp error bounds without explicit compactness constraints on the feasible set or external regularizers. To this end, we derive minimax lower bounds under the standard bounded dynamic range condition, and find that the same Fisher-information geometry in the classic Cramér-Rao lower bounds underpins the finite-sample difficulty of the estimation problem. By combining a non-asymptotic expansion of the likelihood score equation with a fixed-point localization argument, we identify a design-dependent sample size threshold above which the unconstrained canonical MLE exists and is unique with high probability. The same expansion yields a decomposition of the estimation error into a linear stochastic term, an explicit second-order bias, and a higher-order remainder. A refined analysis gives sufficient sample size conditions under which the canonical MLE attains the minimax rates up to logarithmic and constant factors. These results provide a unified non-asymptotic theory for parametric utility elicitation and reveal when the inference is determined by response data alone rather than by external regularization. Preliminary numerical results are consistent with the theoretical findings.

Camera-based 3D object detection and online vectorized HD mapping provide compact scene representations for autonomous driving, but both depend on accurate metric geometry and remain limited by depth ambiguity. Over long-term deployment, observations from repeated traversals can be accumulated into persistent point cloud priors that provide geometric context beyond the current observations. Existing explicit point cloud prior approaches, however, rely on LiDAR-based map construction and therefore require expensive 3D ranging sensors. We propose a framework that constructs a static point cloud prior map from previous camera traversals using Pi3X and augments each point with DINOv3 features. At runtime, a local prior patch is retrieved using global localization, encoded with a sparse voxel backbone, and fused in bird's-eye view (BEV) with lifted multi-view camera features. Task-specific sparse transformer heads then predict 3D objects and vectorized map elements from the fused representation. On Argoverse 2, the vision-based prior improves a strong baseline from 0.287 to 0.299 CDS and from 0.669 to 0.750 vectorized mapping mAP. Ablations show that semantic DINOv3 features are particularly important for vectorized mapping. These results demonstrate that vision-built geometric-semantic priors provide an effective form of long-term scene memory for camera-based perception, improving both tasks without LiDAR for prior-map construction or online inference.

This paper presents a controlled post-alert incident orchestration and response subsystem for educational information systems. The architecture separates deterministic classification, contextual analysis, human approval, and technical execution. A Rule Engine determines severity and selects the playbook, while Static RAG and a local large language model provide advisory content under Validator, Guardrail, Output Sanitizer, and Safe Fallback controls. Experiments begin after simulated alerts are stored in Elasticsearch. The Rule Engine matched the predefined routing matrix in all 30 boundary cases. The Durable Queue completed 100 events without duplicate tasks, new failed tasks, or unintended firewall rules. An eight-alert contention experiment preserved the configured limit of one active model request, and 30 sequential measurements showed an overall mean post-alert processing time of approximately 33 seconds. The results demonstrate functional correctness, traceability, controlled recovery, and bounded model integration within the evaluated laboratory scope.

Orchard robots need maps that preserve small but semantically important structures such as trunks, trellises, and fruit. 3D Gaussian Splatting (3DGS) SLAM achieves high photometric fidelity. However, its optimization remains appearance-driven, and transferring image semantics to 3D points is unreliable for thin structures, whose pixels may receive depth from background surfaces. We present ArborSplat, an online semantic 3DGS SLAM system that tracks with LiDAR odometry and optimizes semantics directly on the Gaussian map, constrained by class-specific height bands above a ground plane fitted to each keyframe's stereo point cloud, and fuses multi-view evidence into a semantic point cloud online while rejecting labels inconsistent with the local ground surface or with monocular depth. Class-constrained refinement reserves Gaussian capacity for underrepresented structures and, under reduced budgets, increases training-view accuracy on tree classes. We evaluate the approach on apple and pear orchards during dormancy, flowering, and harvesting. On full routes, it keeps ATE below 0.5 m on all 12 traversals. On shared 301-frame segments, it exceeds SGS-SLAM and GS3LAM by 0.23 to 0.50 training-view and 0.15 to 0.36 held-out mIoU while running 1.7 to 7.5 times faster, whereas SemGauss-SLAM runs out of GPU memory on all six.

Embodied world models predict the outcomes of robot actions to support learning and planning. For robots equipped with head and wrist cameras, this requires complementary views: the head view captures the overall task, while wrist views reveal local gripper-object interactions. However, evaluating these views independently cannot determine whether they describe the same action and object state. We introduce TRIWORLDBENCH, a benchmark for evaluating embodied world models through synchronized head, left-wrist, and right-wrist videos. It contains 500 episodes across 50 bimanual manipulation tasks and uses 19 metrics to assess tri-view consistency, task alignment, physical and 3D coherence, motion quality, temporal consistency, and visual quality. By combining cross-view checks with measurements tailored to each camera, the benchmark evaluates whether plausible individual videos also form a consistent prediction of the intended task. We summarize overall performance with TWB-Score and retain per-view results to identify where predictions fail. This extends world-model evaluation beyond single-view visual quality. Code, data, and metric definitions are available at https://github.com/TriWorldBench/TriWorldBench.

Recent vision-language-action (VLA) models are promising for general-purpose manipulation, but long-horizon execution remains fragile. Small state-estimation or control errors can lead to irreversible failures (e.g., collisions and object drops). Avoiding these risks requires a proactive safety mechanism capable of anticipating hazards. In this paper, we introduce SafeLoop, a non-invasive external wrapper that adds hazard prediction and rollback-based recovery to a VLA model without changing its parameters. SafeLoop trains a risk predictor from vision and proprioception to output four values: the probability and time-to-hazard for body collisions and for object failures. A lightweight controller then chooses one of three actions based on the predicted risk: continue execution (noop), save a safety checkpoint (record), or retreat in joint space (rollback). Rollback moves the robot back to a recent safe waypoint and queries the base policy again, which may yield an alternative continuation. Across 24 LIBERO tasks (16 random seeds each) and three real-robot tasks (25 rollouts each), SafeLoop achieves a stronger overall safety-success trade-off than alternative methods, reducing hazard cases by roughly 70% while preserving task success and the base-policy control rate. Project code is available at https://github.com/Loule0-0/SafeLoop/tree/release/safeloop.

Well-structured transition systems (WSTS) provide a classical framework for the verification of infinite-state systems, but their probabilistic extensions lack a unified treatment of quantitative coverability: path-enumeration algorithms assume a finite branching degree, while alternative approximation schemes defer some computations, such as probabilities over a bounded horizon, to the model at hand. We introduce probabilistic well-structured transition systems (pWSTS), Markov chains over countable state sets whose underlying transition systems are WSTS, with no a priori assumption on the branching degree. This class encompasses any WSTS equipped with a Markov kernel, such as probabilistic vector addition systems (pVAS) and probabilistic lossy channel systems (pLCS). For an effective subclass, we solve the approximate quantitative coverability problem over bounded horizons, and over infinite horizons under decisiveness, requiring no probabilistic information beyond individual transition probabilities. We then identify a general source of decisiveness: every stochastically monotone pWSTS is decisive with respect to every upward-closed set. We finally instantiate the framework on multi-type Galton–Watson processes, a classical model of population dynamics whose offspring distributions may have infinite support. Under mild assumptions on the reproduction laws, these processes are effective pWSTS, and they are stochastically monotone, hence decisive. Approximate quantitative coverability is therefore computable for them over both horizons, with a proof that uses none of the traditional tools: neither generating functions nor any case distinction between regimes.

In special relativity, a particle can experience constant acceleration, but at the same time, its motion is constrained by the velocity limit imposed by the speed of light $c$. Inspired by this principle, we propose a method for enforcing velocity bounds in control systems by replacing $c$ with the maximum attainable speed of the system. We refer to this as the “Minkowski wrap," the operation of deforming the phase portrait of a system so as to enforce desired speed limits. We apply this idea to shape the input generated by a state-feedback stabilizing controller and time-optimal controller considering controlling a double integrator system. By applying Pontryagin's Maximum Principle, we show that the time-optimal control of a wrapped double-integrator system is bang-bang. The proposed method transforms the classical double-integrator dynamics into a “wrapped" system that respects velocity bounds without the need for clipping, offering an explicit nonlinear feedback control strategy conducive to safety applications.

Graph neural networks have achieved strong performance in node classification by aggregating information from graph neighborhoods. However, a single aggregation mechanism may be insufficient to capture diverse structural patterns across graph datasets. Moreover, independently training multiple structural branches can introduce substantial overhead without necessarily producing stable node representations. To address these issues, this paper proposes PreGS, a parameter-transfer-based multi-expert graph neural network framework. PreGS first pretrains a multi-head graph attention network (GAT) and transfers the linear transformation weights of its first-layer attention heads to multiple GraphSAGE experts. The transferred experts are frozen and used as complementary structural branches. The fused raw node features, GAT head representations, and GraphSAGE expert representations are fed into a multilayer perceptron (MLP), whose output is further fused with the pretrained GAT logits. Based on PreGS, we further develop PreGSv2, which introduces source-level weighting and a structural gating mechanism for adaptive multi-source feature integration. Experiments on eight public graph datasets show that PreGS and PreGSv2 achieve competitive performance against representative graph neural network baselines. Ablation, parameter-transfer, sensitivity, aggregator, visualization, and training-time analyses further validate the effectiveness and stability of the proposed framework. The code and datasets are available at https://github.com/LH-Czc/PreGS.

In the binary diameter problem, the input consists of explicitly given permutations generating a finite group $G$ and a binary-encoded number $k$, and the question is whether every element of $G$ is a product of at most $k$ of the given generators. In the binary length problem, the input additionally contains a permutation $g \in G$, and the question is whether $g$ is such a product. We show that the binary diameter problem is PSPACE-complete, thereby confirming a conjecture of Lohrey and Rosowski (2023). Moreover, we show that PSPACE-hardness already holds for solvable groups of derived length at most 3, whereas the problem is $\Pi_2^{\mathrm{P}}$-complete for 2-step nilpotent groups. For the binary length problem, we establish a sharp complexity threshold: the problem is NP-complete for 2-step nilpotent groups, which extends a result of Lohrey and Rosowski for abelian groups, and PSPACE-complete for groups that are simultaneously 3-step nilpotent and metabelian, which strengthens the main result of Jerrum (1985). Finally, we show that the unary diameter problem, where $k$ is encoded in unary, is $\Pi_2^{\mathrm{P}}$-complete for generating sets of the full symmetric group, confirming another conjecture of Lohrey and Rosowski.

This work presents the task and results of the CHiME-9 challenge for Enhancing Conversations to address Hearing Impairment. The challenge considers the scenario of four-party conversations in a noisy, cafeteria-style environment with interfering speech sources and sound effects. Participants are provided with audio recordings made with Meta Aria glasses and hearing aid microphones, and clean speech samples of the conversation participants. The task is to extract the speech of the conversation partners from the noisy multi-channel recordings with the goal of improving the intelligibility and quality of the speech, evaluated using objective metrics and subjective listening tests. This paper reviews submissions from seven teams and ranks them on a combination of subjective intelligibility and quality. Results show that while the objective metrics do not reflect listener performance, the top systems were able to make substantial improvements over the challenge baseline in both intelligibility and quality ratings.

Stateless UTXO-style execution validates transactions from local and referenced data, supporting parallel validation and predictable serialized-size/weight accounting, but multi-step workflows must explicitly thread state through outputs. However, a prepared next-step transaction may become stale when another valid spend confirms first, shifting consistency maintenance, off-chain tracking, and transaction rebuilding to the protocol boundary and potentially increasing coordination cost and latency. Explicitly addressing this gap, recursive invariants (RIs) provide a transaction-level logic and toolchain in which workflow rules are predicates over a transaction's inputs and indexed successor positions referenced by the RI. Realizing such a successor causes the accepted transaction to re-check its predecessor's RI one step later, carrying the workflow rule forward without application-level shared mutable state or executable logic attached to outputs; repeated one-step checks thereby preserve validation-time locality and make validation work explicitly accountable. Many successor clauses are not decidable when the current transaction is validated, so our small statically typed DSL uses Kleene-style three-valued semantics over true, false, and unknown to defer future-dependent obligations until they become checkable. Alongside the DSL, we formalize UTXO validation and ledger extension in our model, identify the validation-time-evaluable one-step fragment, prove the deduction system sound for the three-valued semantics, and give corresponding transaction-validation and ledger-extension algorithms. Notably, a prototype RI interpreter and benchmarking toolchain evaluate six practice-motivated workflows; the reported traces show approximately linear cumulative validation-cost proxy growth while illustrating staged constraints without committing each step to a preconstructed successor transaction.

Tendon-driven continuum manipulators are widely used in medical applications, where accurate tip-position estimation is essential for precise navigation and instrument positioning. However, patient anatomy and procedural setup impose task-dependent unknown shaft configurations, while friction, slack, and compliance introduce hysteresis, making tip estimation challenging. This paper presents a motor-history-conditioned gated recurrent unit (GRU) residual estimator for three-dimensional catheter tip estimation without direct shaft-configuration sensing. First, an initial multidirectional sweep strategy is applied to calibrate a geometric catheter model backbone, and encode the motor-angle and drive-torque response into a shaft-configuration context vector. During subsequent motion, the context conditions a GRU that predicts a task-space residual correcting this backbone, relying on motor measurements alone. The context remains fixed for the current shaft configuration, while the recurrent state captures the evolving actuation history. Across four disposable intra-cardiac echocardiography catheters and 16 bent shaft configurations, the method achieves 3.3mm open-loop tip RMSE, a 59% reduction relative to the constant-curvature baseline.

Forecasters often score the same units per date against one standardized realized outcome. We show that every standardized forecast splits exactly into a component aligned with this common target and a component uncorrelated with it. Three consequences follow: forecast-error correlation largely mirrors forecast correlation and is therefore a poor measure of diversity; an equally weighted combination beats a no-information forecast only when average alignment is large relative to the combination's dispersion; and the gain from adding a forecaster separates into genuine improvement and mere dilution, which equal-weight admission can mistakenly reward. We develop a cautious selection rule, study it in simulations, and apply it to language-model forecasts of US equity rankings and mechanical signals ranking exchange-traded funds. Selection removes most dilution losses, but no combination beats the no-information forecast.

Despite their strong performance, large language models (LLMs) are bottlenecked by KV cache memory traffic during long-context inference. Sparse attention is widely used to accelerate LLM inference by computing exact attention over a selected subset of tokens. To recover the contribution of tokens excluded from exact attention, recent methods apply coarse-grained compensation to the omitted attention tail. However, existing methods typically select tokens based on attention mass and only then compensate for the unselected tokens. This decoupled design overlooks their interaction: selection should prioritize tokens that would leave the largest compensation error if omitted. To address this limitation, we introduce CompKV, the first compensation-aware sparse attention framework that divides tokens into blocks and explicitly optimizes selection for the downstream compensation mechanism. Our theoretical analysis shows that the residual left by block-level mean compensation is governed by both block attention mass and within-block logit variation. We approximate this residual using compact block-level statistics, yielding a deployable selection criterion. We further develop an efficient asynchronous implementation. Experiments on RULER and LongBench-Pro show that CompKV performs best among the evaluated sparse baselines while delivering up to a $6.85\times$ self-attention speedup over full attention.