Papers

We consider approximate message passing (AMP) with discontinuous denoisers, such as hard slicers, quantizers and hard thresholding, for which the classical state evolution (SE) analysis does not apply. It is first shown that the standard Onsager coefficient; i.e., the average pointwise derivative of the denoiser, misses a boundary term, namely, the sum over the jumps of the jump size times the density of the denoiser input at the jump. Then, for real independent and identically distributed (i.i.d.) Gaussian matrices, a positive noise variance, a fixed number of iterations and a prior with finite second moment, we transfer the Lipschitz SE theorem directly to denoisers, fixed in advance, that are continuously differentiable with bounded derivative except at finitely many jumps, by bounding the distance between the iterates of AMP and of AMP on Lipschitz ramps of the denoisers through the fraction of inputs that cross a jump. It is thereby proven that AMP follows SE, also for test functions such as symbol errors, if its Onsager coefficient contains the boundary term, evaluated either along SE or, as we propose, at the noise level estimated from the residual, with the density computed from the known prior. Simulation results with 4- and 8-level pulse amplitude modulation (PAM) demonstrate that AMP with a hard slicer, whose standard coefficient is identically zero, fails with this coefficient, while with the proposed one, it follows SE and comes close to Bayes-optimal AMP.

Many organizations adapt large pretrained models to their own tasks by fine-tuning on private data. Several of these parties often hold data for the same task and wish to fine-tune a model together without pooling that data. Federated learning (FL) enables joint fine-tuning, but reconstruction attacks on shared intermediate values (the model or its gradients) remain a privacy risk. A one-shot protocol that exchanges one encrypted contribution exposes no intermediate value. Such a protocol still gives the trained model to every participant, which is not permitted where the model is a regulated or proprietary asset. We present HE-OFT, the first cryptographically secure one-shot federated fine-tuning protocol in which no party receives the trained model. Each client fine-tunes a low-rank adapter and a classifier head on a frozen public backbone and keeps the adapter. The client uploads one encrypted head displacement, which the server combines under multiparty CKKS and never decrypts. A quorum of clients returns only the predicted label to the querier. On four text classification tasks and one vision task, HE-OFT reaches 61 to 79 per cent accuracy, against 20 to 48 per cent for a client training alone. HE-OFT keeps 85 to 96 per cent of the accuracy of a disclosed model. A test-time query takes 443.1 to 1713.1 s on one core, or 56.1 to 255.1 s with level restoration on a GPU. Restoring levels at the server cuts the traffic per query from up to 1.6 GiB to 13.5 MiB.

Large language models are increasingly used to simulate clients for counselor training and psychological counseling research, but reliable simulation requires clients to remain psychologically coherent across extended interactions. Existing role-playing methods largely rely on static profile prompts and may exhibit persona drift, unrealistic cooperativeness, or inconsistent psychological states, communicative actions, and emotions. Existing evaluations also lack a unified testbed for both stable client characteristics and evolving psychological dynamics. We propose MASC, a Multi-Agent Self-Calibration framework with latent construct alignment for consistent client role-playing in psychological counseling. MASC combines construct-guided generation, collaborative refinement, consistency verification, and memory-based revision in a closed calibration loop that detects and corrects inconsistencies as dialogue unfolds. We further introduce CRPC-Bench, a benchmark covering session-level profile information and Big-Five personality traits, as well as turn-level psychological state, communicative action, and emotion expression. CRPC-Bench contains 38 motivational interviewing client profiles augmented with personality and emotion annotations. Experiments show that MASC outperforms existing methods across profile, personality, receptivity, and turn-level consistency, with the heterogeneous configuration achieving the strongest overall performance. MASC and CRPC-Bench provide a unified foundation for developing and evaluating psychologically coherent client simulations for AI-assisted counseling research and training.

We study the computation of mixed volumes of convex sets given as rational subspace sections of bodies described by membership oracles. The sets need not have nonempty interior. For positive multiplicities $\alpha_1+\cdots+\alpha_k=n$, we give a randomized algorithm that computes a relative $\varepsilon$-approximation with probability at least $1-\delta$. Its running time is polynomial in Gurvits' factor $\Gamma_\alpha=(n^n/n!)\prod_i\alpha_i!/\alpha_i^{\alpha_i}$, the input presentation length, and $\varepsilon^{-1}$, and linear in $1+\log\delta^{-1}$, with degree independent of $k$. The running-time bound holds for every choice of the random bits, and zero mixed volumes are detected exactly. We extend the capacity-scaling and sampling method of Narayanan–Roy by using rational projections and a smooth mixed Monge–Ampère representation. In particular, the algorithm runs in randomized polynomial time when the number of distinct convex sets is bounded above by a constant.

Frontier large language models (LLMs) can generate heuristic functions that guide search to achieve state-of-the-art performance in satisficing planning, where any plan is acceptable. However, these heuristics are not guaranteed to be admissible and can lead to suboptimal plans. We introduce LeanPlan, the first planning system that finds optimal plans with LLM-generated heuristics whose admissibility is machine-checked. Given a domain description and training tasks, an agentic loop uses planner feedback to iteratively improve a reusable domain-specific heuristic, its admissibility proof and the required domain assumptions. LeanPlan implements the heuristic, its proof and an efficient planner with machine-checked grounding and search in Lean 4. We evaluate LeanPlan on ten domains from the International Planning Competition and three new domains, using test tasks with up to 57 times as many objects as the training tasks. With GPT-5.6 Sol in the agentic loop, we successfully generate heuristics and admissibility proofs for all these domains. With the resulting heuristics, LeanPlan usually expands fewer states than the state-of-the-art Scorpion planner and solves more tasks overall.

Existing research on user fairness in recommender systems has developed diverse objectives. However, it has paid limited attention to a distinct distributive perspective: whether users' contributions to model learning should be reflected in the recommendation benefits they receive. We argue that, in addition to existing fairness protections, a fair system may account for the alignment between users' estimated contributions and the recommendation performance they receive. Such alignment can incentivize sustained and informative engagement, thereby supporting a sustainable recommendation ecosystem. To this end, we propose Contribution-Performance Fairness, a novel fairness perspective which requires recommendation performance to be aligned with estimated contribution across user groups and to remain equitable among users with comparable contributions within a same group. To instantiate this perspective, we introduce the Contribution-Performance Fair Recommender (CPFR), a framework applicable to different backbone recommenders. CPFR constructs ordered user groups from a training-dependent contribution considering interaction volume, loss alignment, and optimization intensity, and jointly optimizes recommendation accuracy with the two fairness requirements. A game-theoretic analysis shows that such alignment can strengthen contribution incentives and improve system-level recommendation accuracy under voluntary contribution. Experiments on three datasets and three backbone models demonstrate that CPFR achieves a strong accuracy–fairness trade-off under the proposed operational metric.

Sensors are useful not only for understanding the world but also for deciding what to do next. Existing sensor models however largely stop at perception: they recognize states or predict outcomes, leaving actions modeled separately through task-specific and often closed label spaces. We introduce Sensor-Language-Action (SLA) modeling, a framework that connects multimodal sensor observations, natural language, and actions within a unified model. SLA uses language as a semantic interface between sensing and acting, allowing heterogeneous actions to be represented, predicted, and explained while remaining grounded in the underlying sensor evidence. We build a large-scale SLA benchmark consisting of datasets that span more than 116,000 individuals, 79 sensor modalities, and 60 action groups, together with a multi-faceted captioning pipeline that aligns user context, sensor dynamics, and action evidence. Building on this framework, we present OpenSLA, a unified SLA model for hierarchical action prediction, state understanding, and action explanation. Extensive experiments on real-world tasks in clinical prediction, operating rooms, and metabolic health verify its superior performance over the state-of-the-art. OpenSLA also demonstrates intriguing capabilities including language-guided evidence grounding and zero-shot generalization to unseen actions and cohorts.

Large Language Models (LLMs) are widely used to generate code. Although their functional plausibility keeps improving, the generated code often contains security vulnerabilities. The functionality-security gap captures code that passes functional tests but fails security tests. A recent longitudinal study of three model families concluded that LLMs become smarter but not safer, with the only considered open-weight family stagnating. Whether this holds for other (open-weight) families and particularly for compact models remains open. We present a longitudinal study of the gap across 32 LLMs from seven model families (five open-weight), covering three successive releases per family in flagship and compact variants. Using CWEval with 119 tasks in five programming languages and 31 CWEs, we compare trajectories across families, model sizes, and languages. Newer models do become safer in absolute terms, although no family closes the gap. Unlike prior work, we find that openness does not separate the families: every considered open-weight family narrows the gap significantly, while Gemini 3.1 Pro keeps a gap as wide as the one reported for Llama. Compact models usually produce less secure code than their flagship counterparts, with notable exceptions (e.g., Gemini 3.7 Flash). At the CWE level, we confirm persistent weaknesses such as log injection (CWE-117) and HTTP response splitting (CWE-113) and regressions in the newest proprietary models on memory and integer weaknesses, and show that the same CWE carries very different risk across languages. From these results, we derive implications for LLM vendors, researchers, and developers. In particular, developers should assume neither that upgrades improve security nor that proprietary models are more secure; they should rerun security checks after each model change and provide secure APIs in the model's context.

Awkward Array is a widely used library in high-energy physics (HEP) for representing and manipulating nested, variable-length data in Python. Previous CHEP contributions have explored GPU acceleration for Awkward Array, demonstrating the feasibility and performance benefits of CUDA-based backend while also identifying limitations related to irregular data access, fine-grained kernel launches, and composability of operations. In this contribution, we present recent developments that build directly on these earlier efforts by introducing a CUDA execution model for Awkward Array based on the Python CUDA Core Compute Libraries (CCCL). Using CCCL, we eliminate the need for custom CUDA kernels and can instead use a high-level Python interface. The CCCL-based approach also enables fusion of multiple Awkward operations into a reduced number of CUDA kernels, addressing kernel launch overhead observed in earlier GPU implementations. Lazy execution allows expression graphs to be constructed and optimized prior to kernel generation, improving performance for analysis workflows involving jagged arrays, combinatorial operations, and reductions. In contrast to earlier approaches, this design also emphasizes extensibility, allowing user-defined Python code to be incorporated into GPU execution paths with minimal boilerplate and without breaking existing analysis semantics. We present performance studies that demonstrate improvements over previously reported eager GPU execution strategies for representative HEP analysis patterns. These developments extend the GPU capabilities of Awkward Array toward a more composable and sustainable backend, aligned with the needs of Python-based analysis at the HL-LHC and beyond.

We review the notion of representing music using point sets and argue that such representations are better adapted than sequential representations for matching patterns in unvoiced, polyphonic music, such as keyboard music. Musical transformations such as transposition, inversion, diminution, augmentation and retrograde can be modelled by geometric transformations in pitch-time representations that combine translation with scaling parallel to and reflection in the time axis. We identify eight types of geometric pitch-time representation that use chromatic pitch, morphetic pitch, morph or chroma to represent pitch and either onset time or midtime to represent time. We illustrate how these types of representation allow us to characterise different types of musical transformation. For example, by using midtime instead of onset time, we can precisely characterise certain retrograde relationships; and by using morph and chroma pitch representations we can characterise transformations involving octave displacements and duplications. We present the concept of a transformation class and consider the three specific classes, $F_{\mathrm{2STR}}$, $F_{\mathrm{2STRMod7}}$ and $F_{\mathrm{2STRMod12}}$. We introduce the notion of an inter-pattern transformation graph for a set of patterns, $S$, and a transformation class, $F$. Each vertex in such a graph represents a pattern in $S$ and there is an edge in the graph from pattern $P_1$ to pattern $P_2$ if and only if $P_1$ can be mapped onto $P_2$ by a transformation in $F$. We show, with the aid of such graphs, that the musical relationships between the occurrences of the HAYDN theme in Ravel's Menuet sur le nom d'Haydn can be precisely described in terms of transformations in $F_{\mathrm{2STR}}$, $F_{\mathrm{2STRMod7}}$ and $F_{\mathrm{2STRMod12}}$ within the pitch-time representations considered.

Since the early adoption of e-commerce, travel and tourism has been a lab for the design of recommender systems: tools that help travelers choose destinations, flights, accommodations, and combine them into itineraries. Data-driven recommendation techniques, ranging from case-based reasoning to reinforcement learning, have been adapted to travelers' needs. The research community has produced multifaceted prototypes of travel and tourism recommender systems (TTRSs), which are context-dependent, multistakeholder-oriented, and more recently, addressing sustainability issues, such as overtourism. Despite this enduring work, TTRSs are not widespread yet. We argue that three limitations can explain this: outdated and sparse data sets used to train and validate TTRSs, algorithms that prioritize prediction accuracy over domain-specific dimensions such as novelty and contextual relevance, and a failure to address the specific needs of travelers. Targeted incremental research could address these limitations, but a disruptive factor has meanwhile entered the ecosystem of tourism information and commercialization platforms: generative artificial intelligence. According to market research, GenAI applications are becoming the primary entry point for travelers planning their trips. This forces research to rethink how TTRSs should be designed and which core techniques should be integrated. We claim that future TTRSs, in addition to offering personalized information filtering, should become more flexible advisors that support decision making, integrating multiple data types and AI techniques, from data mining to natural language processing. Moreover, they must transparently balance the conflicting goals of travelers, service suppliers, platform owners, and local communities. We then outline research targets for building more effective TTRSs, fruitfully combining old and new recommendation techniques.

NVFP4 is an attractive datatype for large language model (LLM) inference, offering compact storage and native tensor-core acceleration. However, preserving accuracy using NVFP4 requires careful quantization. In this work we develop a novel quantization scheme called Optimized Smoothing and Scaling for NVFP4 (OSFP4). For each linear projection it uses a diagonal smoothing matrix whose entries are optimized to minimize the squared matrix-product quantization error under NVFP4, taking into account the rounding procedure that is used (either round-to-nearest, or GPTQ-style successive interference cancellation). This requires performing joint optimization on the smoothing entries as well as the block scales, which is facilitated by analyzing a multiplicative-dither FP4 quantizer instead of the fixed deterministic one. Experiments show that OSFP4 achieves the highest average accuracy among the evaluated competitors in the corresponding quantization settings, while retaining approximately 94-97% of vendor NVFP4 prefill throughput on the measured workloads. Our code is available in https://github.com/neriahbd/OSFP4

As video capture moves to handheld and edge devices, motion blur from camera shake has become a pervasive degradation that lowers perceptual quality and harms downstream vision tasks. The strongest deblurring networks recover impressive detail, yet they remain computationally heavy and overwhelmingly complex, so their quality comes at a cost that consumer hardware cannot pay in real time. This gap between restoration quality and on-device speed is exactly what makes real-time deblurring difficult. We developed and implemented TSRN-RTVD, an efficient video deblurring system that explicitly reconstructs the underlying camera trajectory during exposure and uses the recovered motion to guide restoration. This approach turns the physical cause of blur into a signal that drives sharpening. Our system runs on a single consumer GPU and restores the video at 30 FPS while reaching 30.08 dB PSNR on the GoPro dataset. We demonstrate TSRN-RTVD on consumer devices with interactive side-by-side visualization of the blurry input and the deblurred output, live throughput, and an on-screen view of the recovered camera trajectory. Demo video is available at https://youtu.be/3alMwVrVALU.

Contextual priors improve neural-to-language decoding by reshaping candidate scores. However, confidence is read from the same reshaped scores, so the errors a prior leaves behind can become more confident with no change in accuracy to reveal it. We study how a prior shapes confidence in speech retrieval on MEG-MASC and MOUS using local decoding scores, a contextual prior combined by additive shallow fusion, and the fused top-two margin as confidence. Among initially incorrect predictions, we find a confidence-ordering reversal: a larger margin makes a repair more likely when the correct candidate starts near the top of the local ranking, but less likely when it starts lower. On MEG-MASC, pooled correctness AUROC is 0.87, yet AUROC separating repairs from residual errors falls from 0.70 at initial ranks 2-3 to 0.39 at ranks 21-50. Errors starting beyond rank 20, inside the reversed region, make up 46.6% of all post-fusion errors. We propose a score-level account: a repair must first close the correct candidate's initial deficit, limiting its final margin, whereas a residual error can build a large margin between two incorrect candidates. A causal intervention that changes only the fusion weight moves the reversal to deeper ranks as predicted. Under a word-level LM prior, it keeps moving after accuracy gain peaks, so a weight chosen for accuracy does not settle confidence. Reading local and prior scores separately improves selective decoding: the decoder answers on 74.5% of windows instead of 56.7%, while 92% of output sets still contain the correct candidate. Confidence after contextual fusion should retain the local and contextual evidence behind each prediction, not just the fused scores. Project website: https://confidencereversal.github.io/; Code: https://github.com/AmadeusFake/NeuDecodingConfReversal

Machine learning classifiers for remote sensing imagery are typically evaluated as though every pixel were an independent sample. Spatial autocorrelation violates this assumption, since neighboring pixels carry redundant information which inflates sample sizes. How many independent samples does a satellite image actually contain? For an $n \times n$ image whose spatial correlation persists over a range of $r$ pixels, the effective sample size is $\Theta(n^2/r^2)$, not $n^2$. We prove this as a finite-sample upper bound for classifiers on spatially correlated data, and show via a matching lower bound that the rate is tight, and no algorithm can do better. We extend the results to images with directional correlation and spatially varying correlation structure. Our result justifies spatial cross-validation since block holdout with separation proportional to the correlation range achieves optimal generalization guarantees, while random holdout can underestimate confidence interval widths by a factor proportional to $r$. We validate the theory on synthetic data and satellite image tiles from three sensors (Landsat 8, Sentinel-2, and Sentinel-1).

In this work, we introduce Crazyflow, an accurate, differentiable simulator built on JAX. By leveraging jit compilation via XLA, Crazyflow unifies physics and control into a single differentiable computation graph, enabling massive parallelization on accelerated hardware without sacrificing modeling accuracy. This architecture achieves order-of-magnitude speedups over existing baselines, capable of training deployable reinforcement learning agents in seconds. To highlight its highly modular design, we demonstrate how easily Crazyflow can be extended by integrating a complete, high-fidelity Ultra-Wideband (UWB) and Inertial Measurement Unit (IMU) simulation pipeline coupled with a full-state Extended Kalman Filter (EKF). This capability allows for massive parallel controller evaluation under realistic, degraded state feedback with minimal impact on GPU throughput. By combining speed, accuracy, and extensibility, Crazyflow serves as a foundational tool for the next generation of aerial robotics research.

Portable mobile-manipulation demonstrations can help alleviate data scarcity for embodied intelligence, but obtaining reliable, low-cost, and robot-free motion supervision from RGB observations remains challenging. Existing approaches often rely on teleoperation or specialized devices equipped with additional sensing hardware, while directly using estimated visual odometry (VO) trajectories can introduce inconsistencies due to accumulated drift and imperfect motion supervision. We present the Visual-Odometry-Conditioned Mobile Manipulation Interface (VOMMI), a portable demonstration collection and learning framework that connects portable RGB demonstrations to vision-language-action (VLA) post-training through offline trajectory reconstruction and online visual-motion conditioning. VOMMI synchronizes body and hand views to capture navigation context and local object interactions without requiring human-robot kinematic correspondence calibration. R2-VO refines offline demonstration trajectories using sparse geometric anchors and produces causal local-motion tokens over multiple prediction horizons for online policy conditioning. An action-group residual adapter incorporates these tokens only into the base branch. Experiments use a 500-trajectory portable for each task, with 75 trajectories held out for RGB-VO evaluation, and 200 robot demonstrations as references. Our policy, post-trained only on portable demonstrations, achieves 18.2% lower base-velocity error than a policy trained with robot-collected demonstrations, while maintaining comparable end-effector translation accuracy. Offline reconstruction reduces absolute trajectory errors for the body and hand streams by 24.6% on average relative to the best evaluated baseline for each stream. The complete system improves the mean success rate by 8.3 percentage points over OpenPI 0.5 across three real-robot tasks.

Multimodal vision-language systems typically fuse image and text embeddings through classical operators such as concatenation, attention, bilinear pooling, or tensor interactions. We propose Quantum Entangled Multimodal Fusion Networks (QEMFN), a hybrid quantum-classical framework that introduces parameterized entanglement as a structured inductive bias for multimodal fusion. Pretrained visual and textual features are projected into compact latent spaces, encoded as angle-parameterized quantum states, processed through intra-modal and paired cross-modal entangling circuits, and measured to produce fused representations for retrieval. Under matched parameter budgets and identical frozen CLIP backbones, QEMFN outperforms classical fusion baselines on COCO-5k and Flickr30k, including multilayer perceptron, tensor fusion, FiLM, cross-attention, compact transformer, and a dequantized paired-topology analogue. An ablation suite isolates the quantum module's contribution from the surrounding classical projections, and quantum-centric analyses report Meyer-Wallach entangling capability, expressibility, gradient variance against barren-plateau bounds, and entropy-performance correlation under controls for training progress alongside an intervention study on the entangling component. QEMFN is executed under shot-based estimation, a noise-modeled fake backend, and a real superconducting device with zero-noise extrapolation. This work does not claim quantum computational advantage; the contribution is the framework together with a controlled empirical and quantum-centric evaluation that positions trainable entanglement as an interpretable, hardware-executable fusion mechanism at scales accessible on contemporary devices.

Learning from experience is essential for LLM agents to adapt to unfamiliar and dynmaic environments. Evaluating this ability is therefore important for understanding how effectively agents acquire and use new knowledge. Existing benchmarks have sought to evaluate this ability, but they primarily evaluate tasks whose rules are provided in the instructions or already familiar to pretrained models, making it difficult to distinguish learning from interactions from reasoning with existing knowledge. To address this, we introduce Learn2Play Bench, a benchmark of newly designed text-based games, whose rules are novel or counterintuitive, requiring agents to acquire knowledge through interaction rather than rely solely on pretrained knowledge. These games provide reproducible feedback and automatic scoring, enabling controlled evaluation of learning across repeated attempts. We also vary game instances to test whether agents can apply what they have learned to new situations. Therefore, we evaluate how backbone models, self-evolving methods, and agent harnesses affect agents' learning ability, revealing three findings: (1) Experience retention: Retaining complete records of actions and feedback can support more effective learning than summarizing these experiences into rules or strategies. (2) Human agent gap: Top-performing human players achieve higher peak scores than the evaluated agents. Human explore more varied strategies, and repeat actions less. (3) Harness matters: With the backbone fixed, changing the harness can improve performance while reducing estimated inference cost. Together, these findings provide insights into how LLM agents learn from experience and suggest directions for future work to improve their learning ability. Project website: https://liushiliushi.github.io/learn2play-bench-website/

We report on an approach to teaching logic to mixed groups of computer science, mathematics, and philosophy students, based on the logico-pluralistic LogiKEy methodology, used for more than a decade in courses, summer schools, and tutorials. LogiKEy uses classical higher-order logic (HOL) as a universal metalogic in which object logics, classical and non-classical alike, are encoded by defining their semantics; through these semantical embeddings a single proof assistant (e.g. Isabelle/HOL), with its automated theorem provers and (counter-)model finders, becomes one environment in which students learn, experiment with, and compare logics. After making the pedagogical case for proof assistants in the logic classroom, we present a graded sequence of classroom examples, each transition motivated by a limitation of the preceding representation, by a need for more explicit modelling resources, or by a new application. A liars-and-truth-tellers puzzle leads from propositional to modal logic; the Wise Men puzzle leads on to dynamic epistemic logic; Boolos's curious inference illustrates what a higher-order meta-logic buys, even for automated proof search; Chisholm's paradox takes the sequence into deontic logic, and from standard to dyadic deontic logic; and Gödel's ontological argument brings it to a research-level metaphysical argument. We then rebut the objection that embedding everything in classical HOL is monism rather than pluralism, reflect on three years of teaching such a course, and sketch the portability of the approach beyond Isabelle.

Fringe Projection Profilometry (FPP) requires precise system characterisation to achieve reliable three-dimensional (3D) reconstructions; however, characterisation accuracy strongly depends on robust checkerboard feature localisation, which can deteriorate under challenging imaging conditions such as lens blur and characterisation target orientations. Existing deep learning-based corner detectors are typically assessed using detection metrics and camera reprojection error alone, without considering their wider impact on projector characterisation, camera-projector stereo characterisation consistency, or overall measurement accuracy. In this work, we introduce a complete FPP characterisation pipeline that incorporates deep learning-based corner detection into the standard camera characterisation workflow. We also characterise the projector by sampling phase values at the centres of the white squares in the characterisation target. Rather than treating corner detection as an isolated task, the proposed framework explicitly analyses how localisation errors propagate throughout the entire FPP characterisation chain. Performance is evaluated using detection metrics (e.g., precision and recall), camera and projector reprojection errors, and the camera and projector stereo characterisation. Across a mixed dataset of clean and degraded images, the camera reprojection error is reduced from 1.237 pixels to 0.259 pixels, while the projector reprojection error is reduced by roughly 50%. Dimensional evaluation of reconstructed artefacts shows improved geometric accuracy compared with those resulting from the conventional pipeline. Overall, the findings indicate increased robustness of system-level characterisation under challenging imaging conditions, thereby enabling more reliable industrial FPP measurements.

Data-driven policing often prioritizes incident records over residents' lived experiences. In the Baltic city of Riga, with a history of distrust and limited community-police engagement, this can further alienate the public. To bridge this gap, we propose a Research through Design (RtD) inquiry into the development of Par drošu Rīgu, a civic tool for community-data-integrated policing. With municipal police, NGOs, and city staff, we ask how RtD enables stakeholder negotiation and which interaction qualities support trust and the use of combined community and incident data. The co-design process included workshops that surfaced divergent notions of safety; material probes designed as boundary objects to negotiate among stakeholders; and a pilot deployment showing how combining quantitative and qualitative data reshapes engagement and trust. Mixed-methods evaluation suggests increased officer-citizen interaction, but frictions in sustaining stakeholder collaboration. We contribute (i) an empirical RtD inquiry with public institutions, (ii) an artifact combining physical and dashboard interactions, and (iii) reflections on interaction design as a boundary-spanning practice for trust and infrastructuring.

For a given data distribution $(X_t)_{t \in \mathbb{N}} \sim Q$ i.i.d., we investigate the hypothesis testing problem: $H_0: Q = P_0$ vs. $H_1: Q = P_1$, for two different model probability distributions $P_0$ and $P_1$. In contrast to the standard setting, where analytic densities $p_0$ and $p_1$ are given, here, we consider the density-free setting, where we only have access to i.i.d. simulations $(Z^0_t)_{t \in \mathbb{N}} \sim P_0$ and $(Z^1_t)_{t \in \mathbb{N}} \sim P_1$. For this simulation-based hypothesis testing setting, we construct an e-test martingale, resulting in a sequential test with anytime-valid type-I error guarantees, approximate growth optimality, geometrically decaying type-II error bounds, and asymptotic power one. Most ingredients used in our constructions are variants of well known concepts. The value of this paper lies in the compact presentation of an effective, anytime-valid solution for the density-free simulation-based sequential hypothesis testing case.

We introduce STRUCTURALCOST, a self-paced reading dataset of 475 participants and 40,800 observations isolating the processing cost of long-distance subject-verb dependency resolution. We replicate a low-powered psycholinguistic finding at NLP scale, namely that human reading times at the main verb increase with dependency length, driven by syntactic embedding beyond linear distance. Different language models – spanning n-gram models, SSMs, and transformers – partially mirror this graded difficulty profile, yet underestimate the integration cost humans incur, with a gap that persists across architectures and model sizes. This suggests these models capture the predictive component of human processing but not the full integration cost that working memory imposes. STRUCTURALCOST provides data needed to drive progress toward evaluating the cognitive plausibility of language models.

The constraint satisfaction problem over a set of relations $\Gamma$ (CSP($\Gamma$)) is the computational problem of deciding if a set of constraints admits at least one solution. The classical complexity for finite-domain CSP($\Gamma$) is settled by the CSP dichotomy theorem: it is tractable if $\Gamma$ satisfies a non-trivial algebraic invariant and is NP-complete otherwise. However, not all these algebraic invariants result in efficient algorithms despite being theoretically tractable. A notable case that generalizes linear equations is that of Maltsev CSPs: an $n$-variable instance with $m$ constraints is solvable in roughly $O(n^8 \cdot m)$ time by Bulatov and Dalmau (SIAM J. Comput. 2006) or $O(n^4 \cdot m)$ time by Dyer and Richerby (SIAM J. Comput. 2013). At the same time, arguably, most "natural" and efficiently usable polynomial-time algorithms rarely exceed a quadratic or cubic time bound. In this paper we revisit Maltsev constraints with this question in mind and find a $O(n^2 \cdot m)$ algorithm (for finite languages, for infinite languages we in addition need to take the total size of the instance into account). The main novel idea is to not attempt to improve the bottleneck in Bulatov and Dalmau (the Fix-Values procedure) but to avoid it altogether with a slightly more refined approach that allows us to search through a smaller space.

Assistants grounded in a small, frequently edited knowledge base can retrieve through tool calls to a live data interface or through vector retrieval-augmented generation (RAG). We compare the two on KyGround, a benchmark of 198 questions drawn from the published records of a Greek–English agricultural platform on Kythera, Greece, with answers verified automatically against the records and each question posed in up to nine forms, including Greek without accents, in capitals and in three Latin-script (Greeklish) schemes. With Claude Haiku 4.5 as router and answer model, a reconstruction of the platform's tool agent answered 71.6% of canonical Greek questions correctly and vector RAG 95.3% (difference $-23.6$ percentage points, 95% CI $-33.1$ to $-15.1$). Letting the router write the vector query changed nothing, and placing the whole knowledge base of about 26,000 tokens in the prompt reached 99.3%. The tool agent's losses arose in retrieval. Its literal searches returned nothing when the router's arguments did not occur verbatim in a record, for example when it transliterated Greek into Latin script or combined words that occur in a record but not as one phrase, and the agent then abstained. Unaccented and capitalised questions cost the tool agent about 20 points and vector RAG at most 2; accent-insensitive search removed this loss, and matching stemmed tokens raised the tool agent to 83.8% on canonical Greek. Greeklish cost both designs about 21 to 32 points. Tool interfaces for community knowledge bases need search that tolerates how users type.

This work studies non-linear function computation broadcast (NFCB), in which a sender with access to $N$ datasets $(X_1,\dots,X_N)$ broadcasts a common message to $K$ users, each possessing side information and requesting a function of the datasets. The goal is to minimize the rate required for asymptotically lossless recovery of all demands. We introduce a broadcast graph that jointly captures the source distribution, side information, and demanded functions. Using Körner's characteristic-graph framework, we develop an achievable scheme for arbitrary $K$, general source distributions, and general finite-field demands, including linear, non-separable, and non-linear functions, without restricting the encoding or decoding operations to be linear. We also present a side-information-assisted independent-set scheme and characterize the optimal graph-based achievable rate for compatible functions. For the converse, we derive a multi-letter response-profile bound that strengthens a basic side-information converse, zero-error and asymptotically lossless clique-entropy bounds based on the operational block broadcast graph, and a genie-aided lower bound. For binary NFCB with $N=K$ and side information $X_i$ at user $i$, we characterize the optimal rate in several special cases and bound the worst-case and average additive gaps between the proposed achievable rate and the genie-aided converse for $K=3$ and $K=4$. Finally, three-user examples with Boolean and linear demands illustrate the proposed bounds and show that non-linear encoding can strictly outperform the best scalar and vector linear schemes.

We ask whether specific attention heads, and more finely specific neurons inside those heads, are responsible for recognizing that a language model's context contains network infrastructure information (a hostname paired with its IP address), and whether that responsibility can be validated causally rather than by correlation alone. At the head level the answer is yes, across five models spanning three architecture families: in every model, a small set of heads (1 to 9 out of 128 to 1152 candidates), found by causal ablation screening and tested for selectivity against matched negative and context-free controls, supports a detector with 99.5–100% held-out accuracy. We then ask whether a head's responsibility concentrates into one neuron or stays spread across its dimensions; this is model-specific. In one model, the top head's signal concentrates into a single neuron, found independently by both a causal intervention and a correlational ranking, which agree exactly (AUC = 1.000, matching the full head). In another, the single clean head works as a whole (AUC = 1.000) but the best causally ranked neuron inside it does not (AUC = 0.665), so the responsibility there is spread across the head. The remaining three models fall in between. On an independent dataset collected by a different institution (reverse-DNS records rather than the discovery data), every model's full-head detector flags 100% of positive records; the single-neuron versions transfer less reliably, and in one model score below chance. Causal head-finding for a specific network-information entity works across models and architectures; how far that finding can be pushed down to individual neurons varies, and needs to be checked for each model.

Consider the $\mathbb Z_2$ synchronization problem \[ \boldsymbol{Y} = \frac{\lambda}{\sqrt n} \theta\theta^{\top} + \boldsymbol{Z}, \] where $\theta$ is uniform on $\{-1,1\}^n$ and $\boldsymbol{Z}$ is an independent Gaussian Wigner matrix with off-diagonal variance one. We give a polynomial-time posterior sampling algorithm for every fixed $\lambda>1$, for which the conditional output law converges to the posterior in total variation, in expectation over the observation. The construction combines sequential TAP proposals with an independence Metropolis correction. The key is to control signed overlaps after logarithmic pinning and conditional TAP approximations along a random revealing path, which give an efficiently evaluable proposal with a polynomial density-ratio bound outside a set of vanishing posterior mass. To the best of our knowledge, this is the first polynomial-time posterior sampler for $\mathbb Z_2$ synchronization with a total-variation guarantee throughout the supercritical regime. For comparison, the diffusion-based sampler of \cite{montanari2023posterior} gives normalized Wasserstein guarantees at sufficiently large fixed signal-to-noise ratio.

We prove that deciding whether there is a surjective homomorphism from an arbitrary finitely presented inverse monoid onto the bicyclic monoid is $\mathsf{NP}$-complete. As part of the proof, we show that an extension of existential Presburger arithmetic which involves greatest common divisors on $n$ arguments is in $\mathsf{NP}$, extending a recent result of Défossez, Haase, Mansutti, and Pérez (SODA 2024).