Peer Revenue Equity in science

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Sunday 30 August

Artificial Intelligence

An Open-Source, Event-Driven Pipeline for Cryptocurrency Market Data: Ingestion, Forecasting, and On-Chain Fraud Detection

Cryptocurrency markets generate high-frequency, multi-source data that is expensive to work with unless a team already has commercial-grade streaming and warehousing infrastructure in place. This paper describes a fully open-source pipeline that reproduces the behavior of a cloud-native, event-driven system – file arrival triggering a message, a message triggering compute – entirely on commodity hardware, using Apache Kafka and a filesystem-watching poller in place of managed cloud triggers. The pipeline partitions historical Gemini exchange data into hourly and minutely files, ingests them asynchronously through two independently grouped Kafka consumers (one for audit logging, one for Spark-triggered ETL), and lands cleaned output in a PostgreSQL warehouse with historical and aggregated schemas plus asset-specific data marts. We use the resulting Bitcoin data mart to compare a seasonal ARIMA model against a single-layer LSTM network for price forecasting, and separately apply Random Forest and Gradient Boosting classifiers, with additional engineered features, to the public Ethereum fraud detection benchmark introduced by Farrugia et al. We report the architecture, the modeling methodology, and the resulting metrics, and we are explicit about the limitations of comparing forecasts issued at different horizons and of evaluating fraud detection on a static, already-labeled dataset.
Read the full abstract

Probability

Cramér transform, half-space depth and threshold phenomena for convex bodies

We study the relationship between the Cramér transform and Tukey's half-space depth for log-concave probability measures. For the uniform probability measure $\mu_K$ on a convex body $K\subseteq\mathbb{R}^n$, we prove the sharp pointwise comparison $$\Lambda_K^*(x)\leq -\log q_K(x)\leq \Lambda_K^*(x)+\frac12\log n+C,\qquad x\in\operatorname{int}(K),$$ where $C$ is an absolute constant. The order $\log n$ is optimal, as shown by the Euclidean ball. The proof combines exponential tilting, one-dimensional log-concavity, and self-concordance of the Cramér transform. As consequences, we obtain sharp-order moment and tail estimates for $\Lambda_K^*$ and identify $\exp(\Lambda_K^*(x))$, up to polynomial factors in the dimension, with the number of independent samples needed for $x$ to be captured by their random convex hull. We also establish an $O(n^2)$ variance bound for the logarithmic half-space depth and use it to derive a general criterion for sharp thresholds of random convex hulls. In particular, this criterion applies to the uniform measures on $\ell_p$-balls for every $p>1$. These results establish a quantitative link between large-deviation cost, geometric depth, and sampling complexity in high-dimensional convex geometry.
Read the full abstract

Artificial Intelligence

EDGE: Engine for Deterministic Graph Evaluation through Conversation Simulation from Graph Structured DSL Configuration

As agentic systems evolve into complex multi agent orchestration workflows, there is a growing and critical need for systematic frameworks that measures an agent's behavioral consistency and determinism. In this paper, we introduce a formal evaluation methodology that is grounded in AgentGraph, a planner powered by a domain specific language that represents agent reasoning through a dynamically adjustable directed graph. We leverage this structural formalism and utilize graph traversal algorithms that exhaustively enumerate conversational paths, forming a comprehensive evaluation set that captures the agent's complete behavioral space. We then systematically replay these reproducible trajectories to compare observed outputs and state transitions against the intended DSL specification. To quantify reliability, we define novel metrics that measure response and trajectory determinism, structural adherence and semantic consistency across both exact replays and their linguistic variants. Our system's results demonstrate that agents configured using frameworks like AgentGraph and LangGraph with explicitly structured node transitions show superior determinism over agents that are not configured with controlled transitions.
Read the full abstract

All authors Ram Kulathumani, Pushkar Nagar, Regunathan Radhakrishnan, Anupam Tripathi, Xiangbo Mao, Roshanak Omrani, Keshav Somani, Shwet Kamal Mishra, Shayna Lurya

Computer Vision

Confidence-Aware Ensemble and Long-Word Refinement for Artistic Text Recognition

Artistic Text Recognition (ATR) remains challenging because word images often combine decorative fonts, curved layouts, object-like characters, clutter, and severe distortions. This paper studies WordArt-V1.5 as a standardized benchmark for this setting and evaluates recent scene and artistic text recognizers under a common protocol. We propose a confidence-aware ensemble that combines SVTRv2, PARSeq, and MAERec after fine-tuning on the official training split. The ensemble selects predictions using the minimum confidence over disagreement positions, emphasizing characters that separate competing hypotheses. For long words, where a single character error can invalidate the whole prediction, we add a targeted refinement stage based on Needleman-Wunsch alignment and lexicon-guided correction. On the WordArt-V1.5 Test B split, the proposed system reaches 89.90% Word Recognition Accuracy, improving the best individual fine-tuned model by 1.77 percentage points. The long-word refinement produces a modest global gain, but improves the targeted long-word subset by 2.72 percentage points. Finally, an error analysis of all remaining mistakes shows that 48.8% are associated with labeling issues, visual ambiguity, or illegible samples, highlighting the value of diagnostic reporting for future ATR benchmarks and models. Our source code is available at https://github.com/lucas-azdias/Artistic-Text-Recognition/.
Read the full abstract

All authors Lucas A. Dias, Henrique A. Schulz, Rafaela de Miranda, Guilherme L. Peres, Pedro L. Bittencourt, Rayson Laroca

Quantum Physics

The MS Unidirectional Current as a Generalization of the ABC Absorption Current

For a free scalar particle in one spatial dimension, in a single-pass half-line geometry, we compare the absorbing Robin boundary condition with the generalized Marchewka–Schuss (MS) unidirectional-current family. Robin absorption is characterized by a single parameter $\kappa>0$: once $\kappa$ is specified, its momentum-dependent absorption profile is fixed. In the generalized MS construction, by contrast, the detector response is described by a spectral function $\lambda(k)$, subject to the positivity and subnormalization conditions of the detection law. We show that every Robin choice of $\kappa$ corresponds to the particular MS calibration $\lambda_\kappa(k)=\pi\kappa/(k+\kappa)^2$, which reproduces the complete Robin absorption current for every admissible one-sided spectral amplitude and at every time. Thus the entire one-parameter Robin absorption family is contained within the generalized MS family. The MS construction is more general: the choice $\lambda_{\rm full}(k)=\pi/(4k)$ gives unit absorption efficiency for every wave number, which cannot be achieved by any fixed Robin parameter. In this precise sense, the generalized MS detector process extends the Robin absorbing-boundary family. We also compare their asymptotic behavior in the ballistic fixed-ray regime and in the fixed-distance late-time limit.
Read the full abstract

Number Theory

$D(N)$-quadruples in upper-triangular $2\times2$ integer matrices

We introduce analogues of Diophantine $D(N)$-$m$-tuples in the noncommutative ring $M_2(\mathbb Z)$ of $2\times2$ integer matrices. Besides definitions based on the standard matrix product, we consider a symmetric version defined via the Jordan product $$A\circ B=\frac12(AB+BA).$$ Special attention is devoted to upper-triangular integer matrices $UT_2(\mathbb Z)$, where squares admit a particularly simple description. Motivated by the classical connection between representations of $n$ as a difference of two squares and the existence of $D(n)$-quadruples in commutative rings, we investigate the existence of Jordan $D(N)$-quadruples in $UT_2(\mathbb Z)$.
Read the full abstract

Robotics

What Can One Sentence Fix? A Failure Taxonomy and a Minimal Language Correction for Frozen VLA Policies

When a deployed vision-language-action (VLA) policy fails, fine-tuning it again for every new setting is costly, yet the person watching it can usually say what went wrong in one sentence. We ask what such a sentence can and cannot fix in a frozen policy. Labelling every failing LIBERO-90 task (out of distribution) for OpenVLA-OFT and pi0.5, we find that the failures sort into four modes and that only one, execution misalignment, where the policy reaches the right target but miscalibrates how far or how hard to move, is within reach of a sentence; the other three lack a target or a plan, not a magnitude. For this mode we define the narrowest human correction: one retrospective sentence per task, grounded by an LLM into an action dimension, direction, magnitude term, and time window, becomes a short bias on the frozen policy's actions, after which the policy completes the rest on its own, episode after episode. Re-executed on its task's failed rollouts, it recovers 65 of all 139 in-distribution and 33 of all 328 out-of-distribution failures in simulation; on a real xArm7 after the robot base is moved, it raises pick-and-place from 10/30 to 30/30 in each of two same-object conditions and from 2/30 to 24/30 with a new object, while two long-horizon tasks stay at 0/30 with and without correction, as the taxonomy predicts. A naively prompted VLM that watches the same failures and writes the same kind of correction recovers none in up to 10 attempts, suggesting that the value lies in the person's calibrated sense of magnitude and timing rather than in language itself. Videos: https://correctvla.github.io/
Read the full abstract

All authors Owen Kwon, Pablo Ortega-Kral, Arthur Bucker, Jean Oh

Computation and Language

DataFoundry: Evolving Data Preparators via Recursive Self-Improvement

Domain adaptation of large language models increasingly depends on constructing high-quality training data, yet existing data-preparation pipelines typically address quality only after generation through post-hoc filtering. This creates a fundamental mismatch: data-quality issues often originate from the construction process itself, while quality control is applied only to its outputs. We introduce DataFoundry, a framework for evolving data preparators through recursive self-improvement before large-scale data production. DataFoundry represents a data preparator as an evolvable runtime specification and instantiates its evolution with a Skills-as-Modules architecture, in which a central Controller orchestrates modular skills to compile executable runtimes, diagnose deficiencies on small pilot sets using domain-appropriate criteria, and translate diagnostic feedback into adapters that revise individual preparation components while preserving stable interfaces. We evaluate DataFoundry on DataPrep-Bench across mathematics, finance, law, and medicine, and find that recursively evolved preparators produce training data with higher downstream utility than baselines. Experiments across different backbones further demonstrate that these improvements are not tied to a particular model, while analyses and case studies further reveal the framework's optimization dynamics and illustrate how its evolution unfolds in practice.
Read the full abstract

All authors Cehao Yang, Xiaojun Wu, Xueyuan Lin, Chengjin Xu, Xuhui Jiang, Hui Xiong, Jian Guo

Artificial Intelligence

Review Before Trust: Source-Grounded Integrity Gates for AI-Assisted Personal Health Records

Large language models can convert medical documents into structured data, but plausible output may still be unsupported by the source. Persisting such output in a longitudinal health record, a record that accumulates patient information over time, therefore creates an integrity risk: unverified data may influence later summaries, trends, or preventive-care computations. We introduce an evidence-gated trust-promotion model that keeps generated data provisional until a deterministic monitor verifies it against the source document. The monitor admits a candidate for a specified downstream use only when the source contains a unique supporting quotation, the relevant fields occur within the same laboratory row, and the required provenance is preserved. The generator cannot approve its own output, missing or ambiguous evidence causes refusal, and refused candidates remain available for human review rather than being silently discarded. We implement the model in Medical DataCloud, a personal health-record application, and evaluate it through automated tests and a replay of saved extraction outputs. All 22 conformance and mutation tests pass. The replay covers nine historical laboratory PDF reports containing 102 manually labelled rows. The reports produce 97 numeric candidates: schema validation accepts all 97, an earlier packet-level evidence check accepts 94, and the hardened quotation- and row-level policy admits 72 while retaining 25 for review. The study evaluates system integrity rather than clinical correctness or clinical safety. The results demonstrate the technical feasibility of an enforceable boundary that prevents generated claims from authorizing their own reuse in a longitudinal health record.
Read the full abstract

Exactly Solvable and Integrable Systems

Nijenhuis torsion and Frölicher-Nijenhuis brackets of recursion operators via their full-fledged forms

We present a novel approach to computing the Nijenhuis torsion and Frölicher-Nijenhuis brackets of recursion operators for symmetries, based on their full-fledged forms introduced by Jahnová and Vojčák (2024). In contrast to the conventional approach, which represents recursion operators as maps between shadows of nonlocal symmetries within a given covering, our method allows the Nijenhuis torsion to be computed directly from its defining formula, without requiring any additional mathematical constructions. The same framework can also be used to compute the Frölicher-Nijenhuis bracket of recursion operators and to verify their compatibility directly in their full-fledged forms. The procedure is illustrated by several examples of full-fledged recursion operators for symmetries of differential equations in four independent variables, including a new recursion operator for the four-dimensional universal hierarchy equation that, to the best of our knowledge, has not previously appeared in the literature. A notable feature of this new full-fledged recursion operator is that its shadow component does not appear to admit a reasonable representation in any of the standard conventional forms. Nevertheless, our approach allows us to prove directly that the Nijenhuis torsion of this operator vanishes. For the examples considered, the corresponding Frölicher-Nijenhuis brackets also vanish, confirming the compatibility of the respective pairs of recursion operators. This demonstrates that full-fledged forms provide an effective framework for studying the hereditary and compatibility properties of recursion operators, including highly nonlocal and/or multidimensional cases that are difficult to handle by existing methods.
Read the full abstract

Lattice Field Theory

Symmetric Mass Generation for Domain-Wall Fermions

We report the first numerical examination of symmetric mass generation (SMG) for domain-wall fermions. Our simulation on a two-dimensional Euclidean lattice computes the path-integral of the Fidkowski-Kitaev Majorana chain with two boundaries, corresponding to the two (physical and mirror) walls. We demonstrate by evaluating the lattice Dirac operator eigenvalue spectrum that the local four-Fermi interaction selectively gaps the edge fermions on the mirror wall while leaving the physical wall intact. We further identify a characteristic signature of SMG in the path-integral: the gap of the Dirac operator opens along the imaginary axis of the complex spectrum, in contrast to the standard mass shift in the real direction. The conventional hybrid Monte Carlo algorithm remains practical with reweighting the complex phase of the Pfaffian, which is found to be remarkably small throughout the simulation. These results provide a practical nonperturbative framework for interaction-induced mirror-fermion decoupling of lattice domain-wall fermions.
Read the full abstract

All authors Sho Araki, Hidenori Fukaya, Tetsuya Onogi, Satoshi Yamaguchi

Cellular Automata and Lattice Gases

The istr-graph: Interactive Visualisation of any Classic-Graph in DDLab

Any type of attractor basin (classic-graph) created in DDLab can now be visualised, manipulated, and deconstructed as a drag/drop “interactive state transition graph” (istr-graph). The new istr-graph applies to subtrees, single basins, the basin of attraction field, compression, and all other classic-graph parameters. This is an important update on the pre-existing “interactive basin of attraction field graph” (ibaf-graph) specific to just the complete uncompressed field, but the ibaf-graph is nevertheless retained for some of its unique attributes. These issues are discussed with a focus on the scope and implementation of the new istr-graph.
Read the full abstract

Rings and Algebras

Congruence-simple multiplicatively idempotent semirings and multiplicative divisibility

We resolve two conjectures concerning multiplicatively idempotent semirings. First, we prove that every congruence-simple multiplicatively idempotent semiring is finite, and hence belongs to the finite classification of Kepka, Korbelář and Landsmann. Second, we prove that every finitely generated commutative multiplicatively divisible semiring is multiplicatively idempotent.
Read the full abstract

Representation Theory

Picture groups and green sequences from the perspective of Ringel–Hall algebras

Let $k$ be an algebraically closed field and let $\Lambda$ be a finite-dimensional associative $k$-algebra. We apply Joyce and Brideland's notion of Ringel–Hall algebra to prove results about picture spaces and picture groups. For instance, we construct a faithful group functor for $\Lambda$, i.e. a faithful functor from the $\tau$-cluster morphism category of $\Lambda$ into a groupoid. Consequently, by results of Hanson–Igusa, the picture space of $\Lambda$ is locally CAT(0) provided that the $\tau$-cluster morphism category of $\Lambda$ admits compatibility of last factors. We also show that green sequences are in bijection with certain positive expressions in the picture group, which generalizes a result of Igusa–Todorov beyond hereditary $k$-algebras.
Read the full abstract

Computation and Language

Generative vs. Encoder Models for Multilingual NER: A Comprehensive Empirical Study on Naamapadam

Language is humanity's most consequential technology, yet for over a billion speakers across India's twenty-two constitutionally recognised languages, its digital layer remains structurally incomplete. Named Entity Recognition (NER), the foundational step in transforming raw text into machine-interpretable knowledge, has been studied exhaustively for English but remains largely unsolved across most Indic languages. This paper presents a rigorous comparative study of generative and encoder-based neural architectures for NER on all eleven languages of the Naamapadam benchmark. We evaluate five classic model families spanning sequence-to-sequence transformers and multilingual encoders; four decoder-only large language models (LLMs) fine-tuned with LoRA and 4-bit NF4 quantisation; and nine generative models in zero-to-5-shot inference. Under strict CoNLL span-level evaluation, encoder-based models (mBERT and XLM-R, both F1=0.675 on Hindi) substantially outperform every generative architecture in ten of eleven languages, with gaps of 7.5-40 percentage points against the strongest competitor (Gemma-2-2B: avg F1=0.427). The best few-shot result reaches only 28% of the encoder baseline. We identify three language clusters–encoder-dominant, partial-coverage, and failure-zone; and provide actionable deployment guidelines grounded in transfer learning and low-resource NLP principles.
Read the full abstract

All authors Jakkala Mahesh, Jatavath Shravan Kumar, Komalla Shivani, Sujoy Sarkar

Computer Vision

RIDGE: Region-Informed Derivative-Guided Evidence Selection for Long Video Understanding

Long videos contain far more visual content than Large Vision-Language Models (LVLMs) can process under a fixed visual-token budget, making frame selection essential. Existing query-aware selectors usually estimate frame-query relevance and build a compact subset from high-scoring frames. Although their mechanisms differ, the similarity sequence is still often treated primarily as values to rank or sample from, rather than as an ordered signal whose shape reflects how query-relevant evidence emerges, peaks, and fades over time. This can obscure frames that explain, contextualize, or follow an event, because such evidence may lie on the rising or falling sides of a nearby relevance peak and receive lower absolute scores. We propose RIDGE, a frame selection framework that reads the frame-query similarity curve as a temporal signal. By using local changes and curvature, RIDGE partitions the timeline into structural regions and applies region-specific selection to preserve event cores, transitions, buildup, aftermath, and contextual frames under a fixed budget. It is a lightweight post-processing step on precomputed frame-query scores and requires neither training nor iterative LVLM calls. Across four long-video benchmarks and three backbones, RIDGE achieves the best performance in most settings and remains competitive in the others.
Read the full abstract

All authors Shanqing Xu, Meng Luo, Mengchen Qian, Yuhui Gao, Siyue Peng, Xiaohan Zhong, Xiaojin Zhang, Zhongyu Wei, Wei Chen, Xiang Bai

Probability

Equilibrium fluctuations of the weakly asymmetric inclusion process

We prove that the stationary density fluctuation field of the one-dimensional weakly asymmetric simple inclusion process converges to the energy solution of the stochastic Burgers equation. The key technical ingredient is a second-order Boltzmann-Gibbs principle tailored to the microscopic current. As a consequence, we also obtain Ornstein-Uhlenbeck fluctuations for the symmetric inclusion process in a weakly condensing regime, a result of independent interest.
Read the full abstract

Computation and Language

Detecting Hidden Chain-of-Thought in Large Language Models with Linguistic, Behavioral, and Mechanistic Indicators

Large language models often answer complex reasoning questions without revealing intermediate steps, raising whether they reason latently or complete patterns. We propose the Hidden CoT Detection Score (HCDS), a comparative behavioral and mechanistic signal measuring whether neutral-prompt behavior aligns more closely with explicit CoT or explicit no- CoT. Here, hidden CoT operationally denotes this neutral-prompt CoT-like alignment; HCDS does not directly observe or prove an unexposed reasoning trace. On GSM8K, HCDS is significantly positive for both Qwen3-4B variants (Thinking $+1.87$, $p = 1.2 \times 10^{-7}$; Instruct $+1.41$, $p = 1.9 \times 10^{-4}$), replicates across a different inference stack and quantization within $0.08$ ($+1.80$ and $+1.45$), and is not significantly positive in seven of eight length-adjusted calibration-control cells. The unadjusted score produces large positive scores on single-step arithmetic and numeric factual lookup. The variants also respond differently to no-CoT instructions: Instruct complies from the prompt alone, whereas Thinking continues reasoning and requires intervention. These findings show stronger, less prompt-conditional CoT-like behavior in the reasoning-tuned model, consistent with but not proof of latent reasoning. HCDS thus investigates latent reasoning without relying on models' self-reported traces.
Read the full abstract

All authors Armaan Singh, Ryan Trinh Le, Jasmine Kaur, Abdullah Sultan, Edward Lue Chee Lip, Kiran Nijjer, Adnan Ahmed, Vasu Sharma

Human-Computer Interaction

A Cyber-Physical Machine Tool Framework with a Real-Time Machining Process Digital Twin

Digital Twins (DTs) have emerged as a key technology for improving the monitoring, optimization, and automation of manufacturing systems. However, existing Cyber-Physical Machine Tool (CPMT) implementations primarily represent the machine tool, while the machining process remains only partially synchronized with its physical counterpart. This paper extends a previously presented CPMT framework by introducing a hierarchical DT framework that simultaneously maintains DTs of both the machine tool and the machining process. The proposed framework integrates real-time CNC operational data, a voxel-based workpiece representation, synchronized process vibration measurements, and a persistent part DT repository for process replay, traceability, and future synthetic data generation. Experimental evaluation demonstrated real-time operation at a 20 Hz machining-state update rate, interactive visualization exceeding 100 frames per second, and a mean depth reconstruction error of 0.16 mm. The implementation provides a foundation for AI-assisted machining applications while preserving the machine tool monitoring and teleoperation capabilities.
Read the full abstract

All authors Khalil Chakal, Tero Kaarlela, Jose Outeiro, Carlos Andrade

Statistical Methodology

Modelling compositional data with structural zero values

Compositional data are positive multivariate data whose sum equals 1. A popular method to analyze such data is via log–ratio transformations, which are however not applicable when zero values are present. In this paper we present a conditional logistic normal distribution suitable for compositional data with structural zero values. The model is applicable to arbitrary dimensions and the EM algorithm guarantees a fast implementation. The model accepts predictors and we provide a residual analysis. We further provide principal component analysis and biplots for visualization of compositional data with zero values present. Simulation studies and real data examples illustrate the performance of the model, with and without covariates, and compare it to the Dirichlet analogue.
Read the full abstract

Artificial Intelligence

SearchWiki: Learning to Build and Navigate Knowledge Wikis for Active Information Seeking

Flat retrieval-augmented generation treats a corpus as a bag of chunks, discarding document hierarchy and cross document structure. We introduce SearchWiki, a harness framework that synthesizes a corpus into a hierarchical, typed, navigable wiki and trains an agent, WikiResearcher-9B, to retrieve information through multi-turn tool use. The wiki organizes knowledge into three layers - document overviews, cross- document topic pages, and page-level source records; enabling progressive refinement of retrieval when initial lookup misses. We optimize the agent's navigation policy with on-policy reinforcement learning with a multi-component reward function balancing answer correctness, retrieval quality and trajectory efficiency. Evaluation on ViDoRe-V3 (8 domains), FinanceBench, and memory benchmarks (LoCoMo, LongMemEval, PersonaMem-v2) shows that WikiResearcher- 9B which is our RL-tuned Qwen 9B model, significantly outperforms same-size untrained baselines and exceeds or matches larger external models. SearchWiki paired with WikiResearcher-9B demonstrates that learned navigation over structured corpora is a superior alternative to flat retrieval.
Read the full abstract

All authors Guransh Singh, Vishwajeet Kumar, Arkadeep Acharya, Adnan Qidwai, Jaydeep Sen, Sachindra Joshi

Computation and Language

How Prolific Sellers Self-Present: Dissecting the Communication Patterns of 1.6 Million Reverb Listings

The current paper draws on self-presentation theory and warranting theory to evaluate how the language patterns in an online marketplace reflect seller status (i.e., a prolific seller versus an everyday seller). Using 1.6 million musical instrument listings from Reverb.com in search of content, style, and structural differences in seller product descriptions, the evidence suggested prolific sellers tend to focus more on objective and functional aspects of a product (e.g., its features and specifications) and less on subjective characteristics like tone, relative to everyday sellers. Prolific sellers also communicated in a more narrative-like style, which was driven by an elevated use of personal pronouns, and they used longer descriptions than everyday sellers. Therefore, what prolific sellers focus on tends to be quite technical, but how they communicate this information is typical of a story that is told to potential buyers. Implications for self-presentation theory and warranting theory are discussed.
Read the full abstract

Artificial Intelligence

Spatial Matryoshka Training for Multi-Granularity Visual Document Retrieval

Multi-modal late-interaction retrievers achieve strong retrieval on visually rich documents by representing each page as per patch embeddings and matching at the token level. However, this approach incurs high storage costs. Existing compression methods typically fix a single compression level at indexing time, limiting flexibility. We present ColSNAP (Spatial Nested Average Pooling)1, a training method that generates a nested hierarchy of compression levels directly from a backbone's patch grid. By spatially pooling patch embeddings into pro- gressively coarser tiers and training all tiers simultaneously, a single model learns to support retrieval at multiple compression levels without architectural changes. Crucially, a single encoding pass yields every tier, enabling the accuracy-storage trade-off to be configured at indexing time to match avail- able storage budgets, rather than being fixed during training. We demonstrate that models trained using ColSNAP maintain near full-resolution retrieval performance under substantial compression and that ColSNAP transfers effectively across multiple late-interaction backbones, and achieves most of its improvements via a lightweight adaptation stage applied to a pre-trained retriever.
Read the full abstract

All authors Trishan Singha Roy, Arkadeep Acharya, Vishwajeet Kumar, Jaydeep Sen, Sachindra Joshi

Human-Computer Interaction

The Policy Deficit in AI x Social-Emotional Learning Research

As artificial intelligence (AI) is increasingly integrated into social-emotional learning (SEL) initiatives, the need for evidence-based policy has become paramount. We systematically reviewed 65 peer-reviewed papers that examine the intersection of AI and SEL to investigate how these studies articulate policy implications. Our analysis revealed a substantial "policy deficit" in the current AI x SEL literature: nearly three-quarters of the studies did not mention policy implications at all. Using the "WH-question" framework (Who, What, Why, When/Where, and How), we map the policy implications narratives present in the literature and show that they often lack the specificity and actor-oriented guidance required for effective evidence-informed policymaking. We find a significant association between publication venue and policy engagement, suggesting that current academic incentive structures may prioritize technical innovation and pedagogical feasibility over explicit engagement with governance and regulation. This study identifies a "techno-solutionist" trap, where technical potential is foregrounded while the institutional conditions for responsible implementation remain under-specified. We conclude by proposing a shift from "implication-as-afterthought" to "implication-as-methodology" and offer a set of actionable guidelines for researchers, editors, reviewers, and policymakers to bridge the gap between AI innovation and educational governance. Rather than presenting policy as a generic ethical horizon, we argue that AI-SEL studies should systematically specify Who should act, What actions are recommended, Why these actions are needed, When and Where they apply, and How strongly they are framed, thereby strengthening the translation of AI x SEL innovation into educational policy and practice.
Read the full abstract

Systems and Control

Integrated Transmission and Distribution Expansion Planning Considering Customer Actions and Distributed Energy Resources

The continued integration of distributed energy resources (DERs) motivates research into integrated system planning. While existing literature shows that DERs can reduce system costs, it often overlooks customer behaviour, particularly DER adoption driven by cost savings and incentives. System cost allocation can influence these decisions, affecting DER uptake, grid injections, and overall system efficiency. Therefore, integrated system planning should consider both system- and customer-level benefits of DERs. This paper proposes a multi-step framework that combines an integrated planning model with cost allocation and retailer business models. The integrated planning model minimizes total system costs subject to network constraints, while the cost allocation and retailer models capture customer responses to costs and DER opportunities. By coordinating these models, the framework represents how customer behaviour influences planning outcomes. The proposed approach is demonstrated on a 36-bus system, with results showing that incorporating customer decision-making reduces overall system costs, lowers reliance on transmission-connected generation, and supports more realistic system planning.
Read the full abstract

All authors Abhinav Ayri, Haotian Yao, Mostafa Farrokhabadi, Hamidreza Zareipour

Computation and Language

XQDT: eXplainable and Quantitative Data-Text Alignment Metric with Feedback Signals

Evaluating data-text alignment remains challenging: existing metrics often provide limited explanations for the scores, while prompt-based LLM-as-Judge methods can be expensive and unreliable. We present an end-to-end explainable evaluation metric that fine-tunes a language model to identify omitted, extra, incorrect, and correct data units in a data-text pair. These local judgements are aggregated into precision, recall, and F1 scores, providing both fine-grained diagnostic feedback and an interpretable measure of alignment quality. Across benchmarks, our fine-tuned models outperform LLM-as-Judge methods in error prediction and achieve competitive precision, recall, and F1 scores, while maintaining strong correlation with human judgements. Beyond evaluation, our verifier outputs also provide useful feedback signals for downstream correction and refinement, supporting alignment-oriented improvement of data-to-text and text-to-data. Code and resources are available at https://github.com/guihuzhang/xqdt.
Read the full abstract

High Energy Physics (Phenomenology)

Bose-Einstein condensation and dark radiation production from bosonic warm dark matter

We investigate the quantum statistical effects of bosonic warm dark matter (BWDM) in a microscopic model in which dark radiation (DR) is produced through the annihilation process $\chi\chi\rightarrow\phi\phi$ mediated by a Yukawa interaction. By decomposing the BWDM phase-space distribution into thermal and Bose–Einstein condensed components, we derive the thermally averaged annihilation cross sections for the thermal–thermal, thermal–condensate, and condensate–condensate channels, including the Sommerfeld enhancement. We find that, for the relatively small scalar DR masses considered here, the thermal–condensate channel remains subdominant. The transition between the thermal–thermal dominated and condensate–condensate dominated regimes is characterized by a critical condensate fraction $r_{\rm c}$, which depends mainly on the BWDM mass and the scalar DR mass. For condensate fractions above this critical value, the condensate–condensate channel can dominate the present-day annihilation rate. We further constrain the BWDM parameter space using the observational upper bound on the additional effective number of relativistic species, $\Delta N_{\mathrm{eff}}$.
Read the full abstract

Fluid Dynamics

Solute dispersion in magnetically influenced multiphase flow through a porous tube: axial transport and microrotational effects

This study presents a theoretical investigation of generalized solute dispersion in magnetohydrodynamic multiphase tube flow with porous layers. A two-fluid analytical model is developed for applications in biofluid and environmental fluid dynamics. The model comprises a micropolar (non-Newtonian) fluid core representing the rotational behaviour of red blood cells and a Newtonian plasma periphery embedded with Brinkman and Darcy porous structures, corresponding to the glycocalyx and endothelial layers with distinct permeability characteristics. A transverse magnetic field is incorporated to investigate how magnetic-field-induced modifications of the carrier flow influence solute localisation, with potential relevance to magnetic nanoparticle-mediated drug delivery. Using the generalised dispersion framework of Sankarasubramanian & Gill, analytical solutions are derived to investigate how the coupled axial velocity field and associated microrotational dynamics influence solute transport. The analytical predictions are independently validated through Brownian dynamics simulations, demonstrating excellent agreement for the temporal evolution of the zeroth and first transport moments. The results reveal the previously unexplored influence of microrotational dynamics on solute concentration, convection coefficients and effective dispersion, providing new insights into the coupled roles of translational and rotational fluid motion in biofluid transport. This work bridges an important gap in the literature and establishes a generalized theoretical framework linking magnetic fields, micropolar fluids and porous arterial structures for biofluid transport, targeted drug delivery and clinical engineering applications.
Read the full abstract

All authors Sohel Ahmed, Nanda Poddar, Jyotirmoy Rana, Kajal Kumar Mondal, Niall Madden

Optimization and Control

Sharp Singularity-Degree Bounds for Equality-Generated SDP-RLT Relaxations of Binary Programs

Singularity degree is an important measure of semidefinite programming (SDP) degeneracy, but it is generally unavailable a priori from the problem data. We augment the Shor relaxation of nonempty binary sets $\{x\in\{0,1\}^n:Ax=b\}$ with the first-level Reformulation-Linearization Technique (RLT) equations generated by the defining linear equalities. For the resulting equality-generated SDP-RLT relaxation, we determine the exact worst-case singularity degree. If $\operatorname{rank}(A)=m$ and $0<m<n$, then the associated relaxation has singularity degree at most $\min\{m,n-m\}$, and this rank-nullity bound is attained for every possible rank in this range. Consequently, the worst-case singularity degree over this class is $\lfloor n/2\rfloor$ for $n\geq 2$. This is strikingly smaller than the sharp general bound $n$ for feasible SDP systems with matrix variables of order $n+1$ (Sturm, 2000, Example 2). Thus, for individual relaxations, rank and nullity provide an a priori bound on the otherwise inaccessible singularity degree and on the Hölder exponent in error bounds estimating distance to feasibility from constraint residuals.
Read the full abstract

Computer Vision

On the Role of MRI Sequences in Cross-Dataset Generalization for Brain Tumor Segmentation

Brain tumor segmentation in magnetic resonance imaging (MRI) is a critical task for diagnosis and treatment planning. Despite the success of deep learning architectures such as U-Net and its variants, performance degradation across datasets remains a major challenge, particularly under domain shift and limited annotated data. To address this issue, this study systematically evaluates how individual MRI sequences influence model robustness across two well-known datasets. A ResUNet-based framework is employed, where each modality is trained independently to isolate its effect under a controlled cross-dataset evaluation protocol with tumor size stratification, without target-domain training, or with limited domain adaptation. Results show that the T2f/FLAIR sequence achieves the best cross-dataset performance, with Dice scores exceeding 75%. It consistently outperforms other modalities across most tumor size ranges, while multi-sequence training further improves performance. Additionally, even limited target-domain adaptation yields rapid initial gains, reducing the need for extensive annotations and costly retraining. Our source code is publicly available at https://github.com/henrique-zan/brain_tumor_segmentation/.
Read the full abstract

All authors Henrique Zan Grande, João G. Pitol, Lucas B. Schuck, Rafael V. Serenato, Rayson Laroca, Andre Gustavo Hochuli