Papers

In-sensor computing reduces the cost of transmitting high-resolution image data by performing early-stage processing near the sensor. However, the logic chip integrated with a CMOS image sensor (CIS) is tightly constrained in compute and memory, limiting conventional deep neural network partitioning. We present OASIS, a distributed in-sensor vision framework that uses a lightweight encoder to generate compact, task-relevant representations before off-chip transmission. The encoder is trained end-to-end using task, entropy, and reconstruction objectives, while the decoder is used only during training. OASIS supports two complementary deployment paths. The first applies 4-bit quantization and Huffman coding while preserving the spatial structure required by classification and dense-prediction tasks. The second uses Sobol-based hyperdimensional computing (HDC) to transform the encoder latent into a fixed-dimensional binary hypervector for associative-memory classification. For the SwinViT-based VWW model, mapping a $3\times3\times8$ latent to a 64-dimensional hypervector provides an additional $1.77\times$ communication reduction with less than one percentage point of accuracy loss relative to the 128-dimensional configuration, yielding an overall $18{,}816\times$ reduction compared with raw 8-bit image transmission. We implement the digital near-sensor pipeline on an AMD Xilinx Zynq UltraScale+ FPGA and characterize it using direct board-level power measurements and Vivado post-implementation analysis, together with circuit-simulated CIS models and a 7-nm ASIC projection. Across visual wake-word classification, hand tracking, and eye tracking, OASIS reduces total system energy by approximately $2\times$-$4.5\times$ while maintaining competitive accuracy, demonstrating a practical hardware-algorithm co-design path for communication-efficient in-sensor vision.

This work studies subspace embeddings obtained by two normalized real Walsh transforms, two independent sign diagonals, and uniform coordinate sampling without replacement. The main result shows that the prescribed sample size $k=\min\{n,\lceil Cr/\varepsilon^2\rceil\}$, for a universal constant $C$, suffices to preserve all squared norms on each fixed $r$-dimensional subspace within $1\pm\varepsilon$ with probability at least $0.99$. The result holds for every ambient Walsh dimension and all ranks, and answers Problem TR-01 in the Open Problems in Numerical Linear Algebra repository. The proof controls joint entry cumulants of the transformed projection through connected graph contractions and Walsh character identities. These estimates then bound the expected trace of even powers of a product of centered projections. A two-projection decomposition converts this estimate into control of both spectral edges. Bernoulli sampling at arbitrary densities and a deterministic upper bound near full sampling yield the prescribed number of coordinates.

Natural-language descriptions of optimization problems may be incomplete or vague about numerical information that a solver requires, including costs, capacities, demands, bounds, and penalties. A language model can translate the description into code, but when a required value is absent it must either stop or guess. We present SAILOR, a proof-of-concept system that detects such unsupported numerical choices, asks the user targeted follow-up questions, and updates the optimization model before returning a solution. Questions are prioritized using uncertainty and solver-derived estimates of how strongly each missing value affects the current model. We evaluate the pipeline on 1,723 instances from seven masked benchmarks using an idealized simulator that returns ground-truth values. Exact objective-value agreement ranges from 27.0% to 87.6% across datasets, with 1.4–5.7 questions per instance on average. These results establish feasibility under controlled branch-and-reveal feedback; they do not measure performance with human users or general structural model repair. Code is available at: https://github.com/sshaghayeghs/SAILOR.

The CRITICS project addresses science accessibility and literacy by converging advanced Machine Translation (MT) based on Large Language Models (LLMs) with educational technology. By leveraging MT systems specifically optimized for scientific content, educational institutions can provide accurate, culturally relevant translations of scientific materials in students' native languages, ensuring that complex scientific concepts are comprehensible while maintaining technical accuracy. Building on these translations, the project explores the design and evaluation of innovative science teaching-learning proposals grounded in curriculum-aligned teaching-learning. Thus, CRITICS will investigate key components of scientific argumentation and critical thinking practices together with textual feedback aligned with learning objectives and assessment criteria inspired by competence-based evaluation frameworks. CRITICS aims to break down language barriers to accessing cutting-edge research and educational materials currently available only in high-resourced languages, thereby facilitating the democratization of scientific knowledge and fostering critical thinking in science education.

We present BLInD (Blind Learned Intent Distribution), a compact network that maps recent vehicle-state history (e.g. speed, curvature, indicator, and vehicle type) to a top-k distribution of future ego trajectories, with no camera, LiDAR, map, or object-track inputs. We find that vehiclestate history alone is sufficient to learn a useful multimodal distribution over near-term ego trajectories, and its low-latency nature makes it well-suited for safety-critical deployment. We investigate two distribution architectures, autoregressive (AR) and flow-matching, and train on both mixed-platform opensource and Wayve datasets. Both generalize without datasetspecific adaptation; the flow-matching model achieves best topk ADE/FDE of 0.15/0.37 m on Wayve, 0.15/0.36 m on Waymo, and 0.28/0.59 m on nuScenes, with the AR model reaching comparable coverage. Integrating the distributions into an AEB trigger task, a strict all-candidates policy reduces false positives from 1.51% to 0.11% with AR (13.7x reduction, 94.9% TP) and to 0.06% with flow-matching (25.1x reduction, 98.7% TP) compared to a 1-CTRV policy with 100% true positive score. BLInD runs in 0.87 ms with the AR head and 2.9 ms with the flow-matching head on an NVIDIA DRIVE Orin ECU making it compatible with real-time deployment on automotive ECUs. While existing learned distribution models rely on scene context and blind vehicle-state models typically collapse to a single path, BLInD is learned, blind, and cross-domain simultaneously, a combination not demonstrated by prior work. These results show that such a distribution provides a controllable and plausible intent sampling interface for downstream systems, with AEB as one instantiation.

Effective anomaly detection frameworks are a central pillar of the Industry 4.0 paradigm. In this paper, we introduce an Optimal Transport (OT)-based framework for anomaly detection, designed to detect deviations from normal behaviour in time-series sensor data. The OT-based method requires minimal user input and adapts to real-time data without the need for labelled training data. Our method effectively addresses existing limitations related to data labelling, generalisability, and scalability, demonstrating resilience against short-term fluctuations, noise, and data gaps - common challenges in industrial environments. Additionally, our method provides counterfactual explanations improving the auditability of the approach when deployed in industrial settings. The proposed method learns the mapping between normal and observed operating conditions through a sliding reference window that adapts to the dynamicity of the data. We evaluate our approach on three industrial datasets, from shipping, industrial HVAC systems, and publicly available benchmark data. The method was highly effective in identifying anomalies and reducing false positives, outperforming traditional methods, while maintaining computational efficiency and ease of configuration.

Datasets that ship automatically generated feature annotations invite a question rarely asked of them: would a human agree with those labels? This report answers that for the Objective Projection corpus, a Turkish narrative dataset whose scenes carry a per-scene applied_rules field from a rule-based detector over six craft features – two prohibitions (emotion labelling, simile) and four positive techniques (materialized metaphor, micro-focus, temporal anchor, atmosphere contradiction). Three studies are reported. Study 1 ($n = 120$) scores the detector against blind labels from the scheme's own author. Study 2 ($n = 100$, a disjoint scene set) scores the detector plus Gemini 2.5 Flash and Grok against an independent non-expert rater whose labels were locked before any machine ran. Study 2b re-runs the identical protocol with Claude Fable 5 (High) and ChatGPT 5.5. The central result concerns one rule. On materialized metaphor – closest to the methodology's theoretical core – the five machine labellers returned positive rates of $0$, $1$, $40$, $72$ and $78$ out of $100$ scenes, against a human count of $9$. Cohen's $\kappa$ was at or indistinguishable from chance for five of six labellers, across both human references and both scene sets: $0.004$, $0.015$, $0.000$, $0.019$, $0.027$. Raw agreement ranged from $74.7\%$ to $84.5\%$, an artefact of class imbalance rather than a sign of competence. We deliberately do not resolve this into a single story. Two readings survive: the feature is genuinely inferential and beyond current automatic detection, or the rule's definition is not yet operational enough for any rater to apply consistently – including the human. Distinguishing them needs a second independent human rater, which this report does not have and therefore does not claim.

Many attacks against workloads running on multi-tenant systems are multi-step and history-dependent: sequences of innocuous operations whose malicious nature emerges only over an execution trace. Defending against them requires security policies that are stateful, are enforced within the kernel, and have a precise semantics. Currently deployed proposals fail on at least one count: kernel's built-in syscall filtering and classical MAC frameworks are stateless, while current eBPF-based tools express their policies through ad hoc YAML rules that cannot capture temporal relations among events, and whose semantics is defined only by the implementation. We present BPFence, an in-kernel runtime-verification framework that satisfies the properties above. BPFence provides a policy language with a formal semantics that can express temporal relations among events. It also provides a type system that statically distinguishes events the kernel can control from those it can only observe. Every well-typed policy is compiled into a finite-state monitor proved correct with respect to its semantics, and then into eBPF programs that run inside the kernel. We evaluate BPFence on seven case studies drawn from real-world attack patterns, and on a set of micro- and macro-benchmarks to show that the enforcement overhead remains compatible with production deployment.

Legacy database migrations often begin with incomplete or outdated documentation, leaving physical data definition language (DDL) as the principal evidence of data architecture. However, DDL does not fully encode conceptual intent, and model-generated completions can be plausible without being correct. This study proposes and evaluates a provenance-aware, deterministic-first pipeline for reconstructing logical and conceptual data specifications from Oracle-oriented DDL while explicitly separating observed facts, deterministic derivations, and large language model (LLM) suggestions. The pipeline performs DDL investigation, parsing, consolidation, primary-key backfilling, type normalization, and declared relationship-graph construction before optional LLM-assisted enrichment. It preserves source provenance in the deterministic catalog and declared relationship graph, and records inferred primary-key and foreign-key candidates in a separate reviewable overlay. We evaluated the implementation on 249 artifactized schema samples comprising 1,225 SQL files. The pipeline completed 244 samples (97.99%); 52 completed samples contained no extractable DDL. Across completed samples, the deterministic path reconstructed 208 tables and recovered 278 declared foreign-key records; 168 of the reconstructed tables lacked an explicitly parsed primary key before backfilling. LLM enrichment generated 100 foreign-key candidates in 36 samples, but the parent-table admissibility rate was only 17.9% for logical-specification candidates and 17.5% for conceptual-specification candidates. These findings show that the proposed deterministic-first architecture can preserve an auditable structural baseline, quantify observed primary-key and relationship gaps, and prevent model-generated hypotheses from being silently promoted to source-grounded architectural facts.

The Interconnectedness Coefficient (IC) is a bounded semi-local graph-theoretic node measure designed to identify connector vertices between cohesive network regions. Such connector vertices, also referred to as bridging nodes, may mediate between locally cohesive regions even when they are neither hubs nor themselves highly clustered. The IC preferentially assigns high values to weakly clustered focal vertices whose adjacent vertices remain strongly clustered after exclusion of the focal connection. Candidate vertices are required to have degree at least two. The construction is partition-free, uses information within radius two, and requires no predefined community or module partition. The range and extremal properties of the score are derived analytically. Exact graph families isolate its maximal response to fully cohesive branches, its controlled response to a single cohesion defect, its invariance under a cohesion-free hub extension, and a sharp fragmentation threshold. A separate application to a Human Interactome Map reveals pronounced degree-dependent stabilization of IC values near the network's mean clustering level. This behavior follows directly from the multiplicative definition. If focal clustering tends to zero while mean leave-one-out cohesion in the neighborhood stabilizes, the IC converges to that neighborhood-cohesion level. Among the highly ranked IC vertices are proteins with established interface, scaffold, and adaptor roles in molecular complexes. The IC is therefore positioned as a semi-local connector measure for cohesive network regions.

Biological organisms minimize free energy by balancing two competing demands on their internal world model: it must be accurate enough to predict sensory input, yet simple enough to generalize beyond it. Two mechanisms regulate this balance offline: sleep reduces complexity through gradual synaptic downscaling, while stochastic noise attenuates precision, relaxing the constraint sensory input imposes on synaptic reorganization. Engineered Spiking Neural Networks (SNNs) leave this balance unaddressed, favoring instantaneous, noiseless weight normalization instead. This paper investigates the hypothesis that a biologically inspired micro-sleep paradigm, napping – combining proportional weight scaling with continuous stochastic membrane activity – can replicate the stability of normalization while shedding model complexity. We evaluate this in an unsupervised recurrent SNN trained via trace-based spike-timing-dependent plasticity (STDP) on Gabor-preprocessed MNIST. We tune napping across three regularization regimes by sweeping its duration and membrane noise level, then compare the best configuration against weight normalization. Across all three regimes, well-tuned napping matches the accuracy of normalization: accuracy peaks at brief durations and low noise, then declines monotonically as either grows. Clustering diverges, with the strongest geometric separation arising at longer durations and higher noise – the two terms of free energy pulling apart, accuracy rewarding data fit and structure rewarding the simpler representation that gradual, noisy downscaling induces. This gain carries a simulation cost normalization avoids, so napping is most compelling where representational structure, rather than raw classification efficiency, is the priority.

This work studies the stability and convergence of augmented primal-dual dynamics when constraint values are estimated from samples. Unbiased constraint observations can produce a biased augmented multiplier signal, shifting the equilibria of the mean dynamics. For componentwise inequalities, we give a necessary and sufficient condition for preserving the Karush-Kuhn-Tucker (KKT) equilibria and construct a convex example with a locally exponentially stable equilibrium that violates complementarity. To address this bias, constraint values are estimated recursively before forming the augmented multiplier signal. For smooth convex conic problems, a joint energy analysis establishes boundedness of the primal, dual, and estimation states, vanishing estimation error, and almost sure convergence of the primal-dual iterates to a single KKT point under global regularity and bounded conditional second moments. The result allows nonunique solutions and multipliers while keeping the number of samples per iteration fixed. Numerical studies illustrate the predicted equilibrium bias and examine convergence with nonunique KKT points and nonlinear constraints.

How many members must a committee have before its pairwise majorities can express any collective preference? With enough voters every pattern of pairwise majorities, or tournament, can arise, but a committee of $k$ voters produces only some; for odd $k$, let $N(k)$ be the least number of alternatives on which some tournament arises from no $k$ voters. Only $N(3) = 8$ is known; for five voters our companion paper gave $12 \le N(5) \le 38$ and the first explicit example of moderate size, on 43 vertices. We show $13 \le N(5) \le 22$. The upper bound comes from nine tournaments on 22 vertices that five voters cannot produce: eight one-vertex deletions of doubly regular tournaments on 23 vertices, including every deletion of one of them, and one arc reversal. The Paley tournament on 23 vertices, which Bachmeier et al. (2019) left open after six weeks of SAT solving, is beyond five voters, and so is every other doubly regular tournament on 23 vertices. The lower bound comes from an analysis of all tournaments on 12 vertices. Both doubly regular tournaments on 19 vertices, the Paley tournament and one other, arise from five voters, but not with unit margin, that is, not with every majority carried by three votes to two. The method is a search that places one alternative at a time and exploits the tournament's symmetries; four of the negative results are certified by machine-checkable proofs, and all results are reproducible from a public repository.

Minibatch persistency reuses data instead of reading it: rather than drawing a fresh minibatch at every optimizer step, it takes K consecutive steps on the same one. Absorbed into data echoing in 2019, it has carried one objection – that reuse merely imitates a larger learning rate – and no baseline tuned as carefully as the method itself. This paper runs the missing test. A pre-registered study trains a 49M-parameter Transformer on FineWeb-Edu at minibatch size B in {32, 128, 512}, 8 seeds per cell, tuning the learning rate separately for every batch size and every arm, against a reuse-free control that changes the sampling and nothing else. Each headline claim is a cost to reach a fixed loss, read on four axes: optimizer steps, fresh tokens, seconds, and joules at the socket. We then replicate on new seeds and a newer GPU generation, and put the three arms on one schedule in steps. The outcome of our study is that what minibatch reuse buys is neither speed nor energy but data, and only at large minibatch size: at B = 32 it reads more fresh tokens than the baseline, not fewer. On steps, seconds and joules it is at best free; and at B = 512, where it looks best, a registered control cannot separate the effect of reuse from the position on the learning-rate schedule at n = 8 seeds. The technique is therefore worth using where fresh data rather than compute is the binding cost: a corpus that runs out, a pipeline that pays per sample, a stream that cannot be rewound. Where the data can simply be read again, spaced epochs do as well or better.

Persistent Gaussian perturbations have been shown to prevent asymptotic oversmoothing in recurrent Graph Neural Networks (GNNs) by ensuring a positive stationary Dirichlet energy. However, this global energy bound does not guarantee that individual node representations remain distinct at finite depths. In this paper, we provide a complementary finite-time analysis of the same persistent-noise architecture. Let \(d\) denote the representation dimension and \(\sigma\) the noise standard deviation. We first prove an exact second-moment decomposition for the expected squared distance between any two node representations, yielding the universal lower bound \(2\sigma^2 d\) at every positive time step without contraction or stationarity assumptions. More precisely, conditional pairwise distances have a noncentral chi-square characterization: the noncentrality parameter is the deterministic message-passing separation normalized by \(2\sigma^2\). This yields dynamics-aware fixed-time and finite-horizon near-collision bounds that retain information discarded by the central worst-case analysis. The earlier central Gaussian bound is recovered as the worst-case zero-separation case. We additionally prove almost-sure pairwise noncollision, derive a uniform finite-horizon guarantee, and establish permutation equivariance in distribution for the stochastic dynamics and permutation-invariant graph outputs. Our results complement the asymptotic energy analysis of prior work and provide rigorous finite-time guarantees on node-level representation separation.

This chapter explores how machine learning (ML) is transforming aquaculture, with a particular focus on enhancing decision-making processes and improving operational efficiency. The chapter is structured to first introduce the challenges in aquaculture and the role of AI and then provide an overview of ML techniques in the context of aquaculture, followed by applications, emerging trends, future directions, and case studies. The focus is on real-world applications of ML techniques, including Random Forest, Convolutional Neural Networks (CNNs), and Recurrent Neural Networks (RNNs), as well as emerging technologies such as Graph Neural Networks (GNNs) and large language models (LLMs). Key applications include biomass estimation, species recognition, behavioural analysis, and environmental forecasting. The chapter also highlights the synergy between ML and the Internet of Things (IoT) for real-time monitoring and decision support. Ultimately, ML-driven innovations have the potential to revolutionise fish farming, leading to more efficient, sustainable, and productive practices in the aquaculture industry.

Large language models (LLMs) are increasingly used in specialized university courses, but control-systems questions require coordinated terminology, notation, derivations, and stepwise explanations. Direct general-purpose responses may be inconsistently structured and hard to verify. Using exercises and reference solutions from a Linear Control Systems course, we built a supervised fine-tuning dataset of 360 system-user-assistant conversations. We applied LoRA to Qwen2.5-3B-Instruct and Qwen2.5-7B-Instruct. With identical data splits, inference settings, and evaluation protocols, we compared base and fine-tuned models and tested LoRA ranks r=4, 8, and 16. Evaluation used ROUGE, BERTScore, and structured-output features to measure reference-answer similarity and stability of the Solution-Method-Teaching Points format. LoRA improved both similarity and structured-output stability at both sizes. On the current test set, 7B-r16 achieved the highest ROUGE-L (0.4093) and BERTScore-F1 (0.8643), while r=8 offered a better balance between performance and parameter efficiency. Bootstrap resampling showed ROUGE-L gains of 0.0764 [0.0613, 0.0915] for 3B-r16 and 0.0874 [0.0687, 0.1042] for 7B-r16; both intervals exceeded zero, indicating stable textual-similarity improvements on the current test set. These results suggest LoRA can align open-source instruction-tuned models more closely with the language and pedagogical organization of course reference answers. However, the metrics mainly capture textual similarity and formatting consistency, not domain-specific reasoning or mathematical correctness, which require expert assessment and task-specific rubrics.

Energy-Based Models (EBMs) provide a flexible framework for generative modeling by learning an energy landscape that assigns low energy values to realistic samples and higher energies to unlikely observations. Despite their theoretical appeal, training EBMs remains challenging due to the computational cost of Langevin sampling and the difficulty of efficiently exploring the learned data manifold. In this work, we propose a cooperative Energy-Based Model and Autoencoder (EBM-AE) framework that combines energy-based refinement with manifold projection. The proposed approach jointly trains an EBM with a denoising autoencoder and introduces an iterative EBM$\rightarrow$AE$\rightarrow$EBM sampling procedure in which Langevin dynamics and autoencoder projection alternately refine generated samples. Within this framework, the autoencoder acts as a manifold projection operator that regularizes sampling trajectories, while the EBM performs energy-based refinement toward low-energy regions of the learned distribution. Extensive experiments conducted on the MNIST dataset demonstrate that joint EBM-AE training substantially improves generation quality compared with a conventional autoencoder. Beyond unconditional generation, we evaluate the proposed framework on image inpainting tasks involving structured and random masks. The results show that manifold projection provides the majority of the reconstruction capability, whereas the final energy-based refinement becomes increasingly beneficial as the reconstruction problem becomes more challenging. Taken together, the results indicate that combining manifold projection and energy minimization provides an effective and interpretable framework for generation, reconstruction, and out-of-distribution detection, while offering new insights into the complementary roles of energy-based modeling and representation learning.

We present North Small Translate, an open-weight, LLM-based machine translation (MT) model with instruction-following capabilities built on the same foundation as Cohere's Command A Plus, a mixture-of-experts architecture with 25 billion active parameters out of 218 billion total parameters. North Small Translate is trained using difficulty sampling to obtain challenging documents and a five-step training protocol combining supervised fine-tuning, direct preference optimization, and online reinforcement learning. We prioritized throughput through a non-reasoning base model and supplemented with optional agentic capabilities to unlock translation quality gains. North Small Translate is trained to perform MT-related tasks, including post-editing and quality estimation, as well as related tasks such as general instruction following. The model achieves top MT performance across 50 languages in the class of models under 1T parameters, with no need to run expensive reasoning at inference time.

Machine-learning models are commonly developed under an assumption that training and test data are sufficiently complete, balanced, labelled, and drawn from compatible distributions. In practice, one or more of these conditions is often violated. Measurements may be missing or corrupted, rare classes may be poorly represented, supervision may be weak, and the deployment environment may differ from the training environment. These imperfections are usually treated as separate technical problems, although they alter learning through a small number of shared mechanisms: loss of information, biased empirical risk, ambiguous supervision, and unstable representations. This short survey organises representative methods around these mechanisms. It reviews reconstruction and generation, rebalancing and representation calibration, learning with limited supervision, adaptation across domains and modalities, and reliability under distribution change. The discussion highlights the limits of plausible reconstruction, benchmark-specific correction, and adaptation without trustworthy feedback. It concludes with directions for evidence-aware learning, uncertainty-preserving prediction, and evaluation that separates visual plausibility from decision utility.

In this paper, we propose DualSpecSE, a speech enhancement framework that jointly models Mel-spectrogram and complex spectrogram in a dual-path architecture for improved ASR performance and higher-quality speech reconstruction. The Mel branch learns coarse-grained acoustic representations and produces enhanced Mel-spectrograms for direct ASR usage, while the complex branch refines fine-grained spectral details for high-fidelity waveform reconstruction. Built upon the cross-band and narrow-band blocks from CleanMel, DualSpecSE introduces an interaction module and a fusion module to enable effective information exchange between the two branches. The model simultaneously outputs enhanced Mel and complex spectrogram without requiring a pretrained vocoder. Experimental results demonstrate consistent improvements in speech fidelity, perceptual quality, and ASR performance. Codes and audio samples are available.

Continuous-latent Autoregressive Diffusion Transformer (AR-DiT) models have demonstrated immense potential in zero-shot speech generation. However, they still suffer from limited decoding stability when synthesizing long utterances or complex linguistic structures. This instability primarily stems from a restricted historical receptive field and an acoustic inertia dependency within the diffusion decoder, which causes the model to ignore semantic conditions. To address these challenges, we propose DiTAR+, a dual-optimization framework. First, we introduce Dilated Context Sampling to expand the macro-level historical receptive field without violating physical temporal continuity, thereby preventing cumulative error propagation. Second, we propose Hierarchical Acoustic Masking to prevent shallow layers from attending to acoustic pre-context, explicitly decoupling semantic alignment from acoustic detail reconstruction. Extensive experiments show that our framework effectively mitigates pronunciation errors and semantic hallucinations, enhances generation robustness on challenging sentences, and maintains exceptionally high speaker similarity throughout the entirety of long-form utterances. On the linguistically challenging ZH-Hard set, DiTAR+ reduces the word error rate from 12.478% to 9.893%, and on extended utterances of 25 to 35 seconds it improves speaker similarity from 0.741 to 0.759 while simultaneously lowering the word error rate from 2.778% to 2.173%, outperforming both discrete-token and pure flow-matching baselines.

Pharmaceutical sales forecasts inform planning across products, regions, and distribution channels, yet their interpretation depends on inventory availability, transaction semantics, product lifecycle, and the information available when each forecast is issued. A model prediction alone does not preserve these conditions or establish whether a forecast is suitable for operational use. We present Phorecaster365, a human-supervised reference architecture that connects enterprise resource planning data to reviewable pharmaceutical sales forecasts. The architecture separates source ingestion, product-region series construction, temporally valid feature generation, statistical and machine-learning modeling, ensemble formation, uncertainty assessment, planner review, and lifecycle governance. Its central intermediate representation is a forecast context package that preserves the source snapshot, forecast target, temporal cutoff, available covariates, data-quality state, and hierarchy version. A corresponding forecast evidence package links predictions to model and calibration versions, exceptions, human adjustments, and publication history. Development experience with a synthetic panel of 10,950 daily records across 30 product-region series informs the design. Historical experiment summaries are retained only as descriptive evidence of development because their evaluation does not establish independent predictive validity. We specify a rolling-origin evaluation protocol, baseline and ablation requirements, uncertainty and robustness assessments, and a staged pathway from synthetic testing to use in governed planning. The contribution is an implementation-neutral system design and validation framework; the report does not establish real-world forecasting accuracy, comparative superiority, or operational benefit.

We study how to update a pre-built aerial scene with a newly captured, unposed ground-view sequence. The aerial scene already contains a reliable metric Structure-from-Motion (SfM) reconstruction and a pre-trained 3D Gaussian Splatting (3DGS) model, whereas the ground-view sequence is collected later to add street-level appearance but has unknown camera poses and global scale. Registering this sequence to the aerial SfM reconstruction is challenging because single-image cross-view localization is brittle and long trajectories are prone to drift. To address these challenges, we present SkyAnchor, which treats the existing aerial scene as a fixed scaffold for ground-view registration and scene update instead of jointly reconstructing aerial and ground imagery from scratch. It first localizes short groups of consecutive ground frames against geometrically verified aerial support, producing sparse anchor poses. It then recovers the full ground trajectory with anchor-constrained submaps, fixing the front and rear anchor poses during incremental registration and bundle adjustment. Finally, it inserts filtered ground Gaussians while preserving the aerial view, followed by lightweight joint refinement. Experiments on seven real aerial–ground scenes show accurate metric ground trajectories and updated 3D Gaussian scenes with strong aerial- and ground-view rendering quality.

Multi-charger scheduling in wireless rechargeable sensor networks must weigh sensor death risk, charger energy, travel cost, return-to-base feasibility and inter-charger coordination at once, and schedulers driven by local urgency alone duplicate service and leave whole regions unattended. We present HQARRF, a two-level scheduler. Below, an interpretable ARR-F score ranks candidate clusters through an attraction term for local urgency, a repulsion term against charger crowding and a force bonus from nearby critical sensors. Above, adaptive zones compress regional state into a deadline-based risk estimate, and a gated Q-learning controller decides only whether to redirect service to a high-risk, under-served zone. Over 27 parameter points HQARRF attains the highest mean survival rate at 26, improving survival by 20.7 percentage points over the mean of five baselines and 9.2 over the strongest baseline at each point. An ablation isolates the upper level: its gain tracks how often the controller fires.

The BabyLM challenge measures how much language a model can learn from developmentally-plausible, child-scale data rather than internet-scale corpora, yet prior language models forgo the biological constraints of the neural circuitry that acquires human language: spiking neurons separated into excitatory and inhibitory populations wired by a recurrent lateral connectome. This paper presents URCHIN (Unified Recurrent Connectome with Horizontal Integrate-and-fire Neurons), which applies the Parallelized Hierarchical Connectome Spiking State-space Model (PHCSSM) to language modeling: leaky integrate-and-fire neurons coupled by a Dale's-law lateral connectome resolve each token through a multi-transmission loop that recirculates activity to a fixed point. The instantiation is deliberately minimal: a single horizontal layer of 128 neurons, no attention, and 4.23M parameters. Two implementations share one set of weights and produce identical benchmark scores, so URCHIN is trained once and deployed either way with no conversion step: a parallel state-space model (SSM) scan that is GPU-efficient for training, or an event-driven recurrent spiking neural network (RSNN) with constant-cost inference for CPU or neuromorphic edge deployment. Across all three BabyLM tracks (Strict-100M, Strict-Small, and Multilingual), URCHIN offers a biologically plausible, efficient, and directly deployable reference point.

With the emergence of online social networks as a primary platform for advertising and advocacy, enhancing a user's influence has become of significant interest. In this paper, we investigate this problem under the Friedkin Johnsen opinion dynamics model, wherein stubborn agents influence the opinions of other agents in the network. Unlike most of the existing works, we leverage topological properties of the network to increase the influence of a desired stubborn agent. Specifically, we introduce the notion of edge modification, which mimics the mechanism of recommendations in social networks. First, we present a topology-based condition that identifies edge modifications that always increase the influence of a desired stubborn agent. It is shown that the impact of the chosen edge modifications remains robust to changes in parameters such as stubbornness and the interaction weights. Thereafter, we formulate a discrete optimisation problem to identify a set of edge modifications that maximise the agent's influence centrality. We present a computationally efficient approximate solution to the optimisation problem. Finally, we demonstrate the effectiveness of our approach on the Friedkin-Johnsen opinion dynamics over the Erdos Renyi random graph.

Earnings conference calls are a primary channel through which managers disclose information under analyst scrutiny. Prior work has linked vocal and lexical cues to future adverse outcomes, but often pools features over an entire call and underuses the interactive structure of Q&A. We propose a two-branch late-fusion framework for detecting managerial evasiveness as a predictor of extrinsic SEC events (primarily late filings): (i) an LLM-as-a-judge that maps Q&A text to an interpretable call-level vector X_text via a structured binary rubric, and (ii) a frozen conversational encoder whose temporal hidden states are read by a DeepVoice-style sequential reader to produce an audio representation h. Late fusion of (X_text, h) yields a call-level risk score p. On n=1,039 calls (212 late filings) with firm-grouped 5-fold CV, fusion reaches AUROC approx. 0.89, versus 0.55 for the text judge and 0.71 for duration alone. These results show that conversational audio dynamics encode managerial evasiveness beyond lexical content and call length, yielding a stronger early-warning signal of adverse SEC outcomes.

We study the recovery of forward and reverse quadratic optimal-transport maps from unpaired samples in high dimensions. We introduce a bidirectional neural framework in which the learned maps induce forward and backward displacement interpolants, while the training objective combines bidirectional quadratic action, discriminator-restricted Jensen-Shannon endpoint objectives, and two-sided cycle consistency. The construction requires neither precomputed sample pairings nor an explicit convex-potential parameterization. For absolutely continuous probability measures supported on a compact convex set, and under the stated generator-approximation, discriminator-richness, and minimizer-attainment conditions, we prove a population recovery theorem: for every prescribed accuracy, the sum of the corresponding \(L^2\) errors between any global minimizer and the forward and reverse quadratic Brenier maps is below that accuracy, provided the discriminator level is sufficiently large and the annealing action weight becomes sufficiently small. Moreover, the cycle loss is bounded above by \(\lambda W_2^2(\mu_0,\mu_1)\). Complementary results quantify approximate invertibility and show that exact endpoint Jensen-Shannon divergence and cycle consistency control missing target mass and many-to-one map collapse, respectively. Experiments on Swiss roll, MNIST, CelebA, single-cell perturbation data, and chest X-ray images evaluate endpoint fidelity, transport cost, inverse consistency, and the geometry of the induced interpolations.

Multi-agent code-generation systems often use one communication topology for every problem. Across five topologies and 614 problems from APPS, HumanEval+ and LiveCodeBench, the advantage of hierarchical collaboration over a single agent grows from 2.4 points of pass@1 on the easiest third to 21.1 points on the hardest third, at about ten times the token cost. We propose the Difficulty-Aware Topology Selector (DATS), which predicts each topology's success probability and maximises predicted success minus cost. Its graph predictor treats topologies as nodes of a connectivity order. The complete graph-head comparison reports a 1.7-point advantage over a flat multi-label head while changing several mechanisms; capacity-matched adjacency controls show smaller descriptive differences. A scalar cost penalty can be recalibrated without retraining for approximately matched-spend comparisons when targets are attained. With a shared penalty, six methods span 21.6 percentage points of spend; two leading baselines fall behind DATS after method-specific calibration near the common target. At 40% of always-hierarchical cost, DATS reaches 77.7% pass@1 versus 73.6% for always-hierarchical and 74.3% for the strongest learned competitor. The main comparison includes Holm–Bonferroni-adjusted McNemar tests; unequal-cost differences do not imply matched-budget superiority. Backbone-specific gains are 4.07–4.23 points across four backbones. Alternatives to the 39 interpretable features, including a hybrid encoder, yield a largest observed gain of 1.3 points, with no significant pairwise improvement; this establishes neither equivalence nor an upper bound. A study on 400 mathematical reasoning problems reproduces difficulty-dependent collaboration gains, widening from 2.5 to 20.9 points.