Artificial Intelligence
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
Wednesday 12 August
Discovery—formulating novel generalizations—is a central part of the scientific process. Despite its importance, there is a gap in the current AI benchmark landscape, with few benchmarks directly probing the capacity for discovering new knowledge with experimentation in controlled environments where the objective is unknown. To address this gap, we release a new benchmark: DiG-bench (Discovery in Games). DiG-bench consists of a set of 70 independent games. Each game is encoded as a short string and has unique transformation rules that must be discovered through interaction and experimentation. The levels of the game present a series of challenges to test whether the rules have been discovered, where the win conditions for each level are also unknown. We provide games at seven tiers of difficulty for AI agents. The lowest tier is routinely solvable by multiple models, while the highest tier challenges the best models in agentic harnesses. All 70 games were solved by at least one human on first attempt. A subset of 21 games is released publicly, and the remainder is held private for secure evaluation.
Machine Learning
Represent, Then Generate: Multimodal-Conditioned Time-Series Generation under Irregular Missingness
Continuous physiological time series underpin modern clinical monitoring, yet many of the most informative signals are invasive, expensive, or simply unavailable for a given patient. Conditional generation offers a remedy: an absent signal can be synthesized from co-recorded signals and routine clinical variables. Existing generators, however, are built around a single conditioning modality and degrade when forced to handle the heterogeneous, irregularly missing mix of time-variant signals and static covariates seen in practice. We propose ReCoGen (Represent Conditions, then Generate), a two-stage framework that decouples multimodal condition representation from target generation. Stage I trains one masked autoencoder per modality, distilling each time-variant condition into a compact and missingness-tolerant token sequence. Stage II trains a flow-matching generator that fuses these tokens with static conditions to synthesize the target signal. Across three physiological benchmarks, including continuous glucose monitoring on AI-READI and arterial blood pressure generation on MIMIC-III and MIMIC-IV, ReCoGen attains the best downstream utility on all sixteen (dataset, task, metric) settings, surpassing six representative conditional generators; on thirteen of them its utility also reaches or exceeds the utility measured on the real signal, a reference we read as an approximate anchor rather than a ceiling. Ablations trace the gains to the conditioning path: learnable cross-attention over the frozen per-modality encoders, and a dual token-plus-AdaLN route for the static conditions. ReCoGen thus turns routinely collected signals into informative surrogates for invasive or unavailable ones, a step toward less invasive, lower-cost continuous clinical monitoring.
Read the full abstract
Systems and Control
The $\theta$-Symmetric SRG with Applications to Stability of Cactus Dynamic Networks
In this paper, we systematically study a variant of the scaled relative graph (SRG), referred to as the $\theta$-symmetric SRG, and apply it to the stability analysis of cactus networks. Compared with the previous SRG definition, the $\theta$-symmetric SRG enables the characterization of phase lead and lag behaviors, and serves as a more natural multivariable extension of the classical Nyquist plot. We first analyze the gain and phase aspects of $\theta$-symmetric SRG separately and build a connection between $\theta$-segmental phase and a norm minimization problem. This connection makes it possible to compute $\theta$-segmental phase via semidefinite programming. We further derive the submultiplicative and subadditive properties of $\theta$-symmetric SRG. These algebraic properties are crucial to determine the nonsingularity of product-type and sum-type return difference matrices, which topologically correspond to the cyclic and parallel-feedback extreme cases of cactus networks. By taking the cyclic interconnection as the fundamental starting point, we establish necessary and sufficient conditions for its robust stability. Integrating this with the parallel case, we synthesize a unified stability framework for general multi-loop cactus networks. The $\theta$-symmetric SRG framework is less conservative and provides a more intuitive geometric interpretation of system behaviors compared with some existing approaches. Several examples are included to demonstrate the effectiveness of the proposed methods.
Read the full abstract
Artificial Intelligence
Auditable agentic AI for evidence-grounded thyroid ultrasound diagnosis and reporting
Thyroid ultrasound diagnosis requires coordinated lesion localization, measurement, risk stratification and reporting, yet most AI systems address these tasks in isolation and provide limited support for clinical review. We present ThyroidXAgent, a clinician-interactive agentic AI system that coordinates specialized diagnostic tools and stores their outputs as an auditable case-level evidence record. The system was developed using OpenThyroidDB, a multicentre, multitask resource integrating approximately 0.3 million ultrasound images and 24,000 paired reports, and was evaluated on 28,458 non-overlapping test cases, including 8,721 cases from 35 centres in the private NHC-MISD-TUS cohort. Across heterogeneous datasets, ThyroidXAgent achieved a mean Dice score of 87.21 percent for nodule segmentation and a mean AUROC of 0.9466 for benign-malignant classification. The same workflow supported lymph-node metastasis prediction and follicular versus papillary thyroid carcinoma classification, with AUROCs of 0.864 and 0.805, respectively. For report generation, evidence-grounded assembly outperformed multimodal language-model baselines across three cohorts. ThyClinScore, a lesion-level clinical semantic metric introduced here, showed the strongest correlation with a location-aware language-model judge. ThyroidXAgent improved physician classification accuracy, increased report diagnostic consistency from 70.3 percent to 86.2 percent, and reduced segmentation and reporting time by 35.9 percent and 27.4 percent, respectively. These findings support auditable, clinician-correctable agentic AI for thyroid ultrasound diagnosis and reporting.
Read the full abstract
Machine Learning · Cross-listed from q-fin.MF
DYSANOS Generative Dynamic Smooth Arbitrage-free Non-parametric Option Surfaces
This article presents with DYSANOS the first generative market model for smooth SANOS option surfaces for all strikes and expiries which are free of static arbitrage. Our model is designed to generate entire paths of daily spot and option prices for years in the future. We present a robust and useful if somewhat simplistic baseline in the form of an AR(1) model. We discuss model setup, data pipeline, and training and investigate market reconstruction, stability, and tail behavior. We illustrate model performance on 891 Option Metrics IvyDB S&P Index surfaces from 2022-01-03 through to 2025-08-29. We also demonstrate how to construct numerically a risk-neutral density. As part of this we develop a new test for zero conditional means under a given measure. We show that for a set of 100,000 simulated paths a trading universe of 48 options and spot is numerically free of dynamic arbitrage.
Read the full abstract
Artificial Intelligence
Reasoning Jury: Multi-Model Consensus for Evaluating Reasoning Traces
Improving reasoning LLMs requires the ability to judge the quality of long reasoning traces for effective reasoning data curation, strong training signals during reinforcement learning, and an in-depth understanding of reasoning behaviors during model performance evaluation. Additionally, surfacing reasoning mistakes that the model makes would enable improving the model's performance at runtime through providing feedback. Due to the difficulty of this complex task on long reasoning traces, single-model judges (even frontier models) do not do well at identifying reasoning defects. Additionally, leveraging frontier models during online training of reasoning LLMs is generally prohibited due to guardrails in terms of use. In this work, we introduce Reasoning Jury, a system that replaces the single judge with a jury of LLMs and a moderated consensus mechanism, to improve the fidelity of judgments for identifying reasoning defects. In reasoning jury, defects of a reasoning trace and their severity are surfaced through a deliberation where a moderator conducts a discussion amongst the jury where the jurors critique each other's judgments and get to modify their initial votes. The moderator derives a consensus through deliberation amongst jurors or consolidation of judgements. We show that Reasoning Jury with a jury of open-weight models (e.g., gpt-oss-120b) is able to significantly outperform frontier models (opus-4.6, sonnet-4.6, and gemini-3.1-pro) at correctly identifying reasoning defects. Besides accuracy performance improvements, the aggregated cost of the jury (initial verdicts, deliberations, consolidation, etc.) is a fraction (8 to 15%) of the cost of running frontier models in LLM-as-a-judge setup. We also show how these judgements can be leveraged to understand failure modes of reasoning LLMs on benchmarks, which allows much deeper understanding of a model's performance.
Read the full abstract
Human-Computer Interaction
Not All Nudges Land: Behavioral Controllability and Elaboration Quality in AI-Supported Journaling
AI journaling tools can tailor prompts to a person's own sensed behavior, but it is unclear which behaviors respond to them. We analyzed 369 journal entries from an eight-week passive sensing study. An LLM labeled each entry as expressing an intention to change a behavior or not, and we measured follow-through against 26 sensor features with a 3-day before/after comparison. Responsiveness depended most on whether a behavior involves other people. Behaviors that depend on others improved in only 15 to 22% of cases, while behaviors a person can act on alone improved more often, up to 50 to 63%, though unevenly. How users wrote mattered less. No single text feature separated improved from unimproved entries; writing carried signal only within specific behaviors, most clearly for text messaging and for longer, more personal intention entries. The sample is small, so we treat these as exploratory patterns that point to where AI journaling nudges are most likely to work.
Read the full abstract
Social and Information Networks
Hate speech toward migrants on a citizen reporting platform concentrates in neighborhoods undergoing demographic change
Understanding when migration generates social integration or exclusion is a central challenge for urban communities. Existing research has mostly relied on surveys, administrative data, or aggregate indicators that fail to capture expressions of exclusion at fine spatiotemporal scales. Here, we analyze over 550,000 geolocated reports from SOSAFE (Chile's largest citizen reporting platform) to examine the relationship between migration and hate speech in Santiago. We fine-tune a Spanish hate speech classifier and validate it against human labels. Reports that mention migrants are more likely to contain hate speech than other reports. Hate speech concentrates in areas with recent demographic change (post-2010 arrivals) rather than in established migrant communities. The spatial analysis shows that hate speech hotspots coincide with neighborhoods where recent migrants comprise over a third of the population. Coldspots appear in high-education sectors with minimal recent migration. Reports with hate speech and reports that mention migrants receive more engagement, although their combination is not amplified further. These results show digital bordering on a citizen reporting platform: exclusionary discourse concentrates, and receives more engagement, in neighborhoods undergoing recent demographic change.
Read the full abstract
Human-Computer Interaction
InvisIto: Weaving Unobtrusive Infrared Markers for Ubiquitous Textile Interaction
Textiles are increasingly explored as media for interacting with digital information. However, many of the existing approaches rely on visible tags, printed overlays, or electronic modules that compromise the fabric's aesthetic and tactile qualities. To address this, we present InvisIto, a method for weaving visually unobtrusive yet machine-readable infrared markers directly into fabrics using near-infrared (NIR)-absorbing yarns. Although these yarns look similar to standard fibers in ambient light, they produce strong contrast in NIR imaging. Our method includes: (1) a design tool that helps users easily embed infrared markers into weaving drafts, (2) five disguising strategies that further reduce marker visibility under ambient light, and (3) a camera-based detection pipeline for decoding and tracking the woven markers. InvisIto supports both woven QR codes for data encoding and woven ArUco markers for binary input and deformation tracking. We demonstrate applications across hand weaving, Jacquard weaving, and industrial fabrication, showing that InvisIto supports scalable interaction and fabrication from bespoke artifacts to mass production.
Read the full abstract
Game Theory
A Repeated-Game Framework for Incentives in Decentralized Infrastructure Protocols
We introduce a repeated dynamic incentive framework for characterizing when "compliance", or full-effort honest service provision, is incentive compatible in Decentralized Physical Infrastructure Networks (DePIN). We model quality control in this setting as a repeated moral-hazard problem between the protocol and each provider, where compliance is enforced by both slashing posted collateral and the discounted threat of demotion to probation tiers on a reputation ladder. Our main contribution is the "deterrence ratio" $\Gamma$, the worst-case ratio of a deviation's private gain to its marginal probability of detection. We find that if any profitable deviation does not increase the fail probability relative to compliance, then binary public-outcome protocols cannot deter it. When all profitable deviations have positive detection gaps and no weakly costlier action is less likely to fail, compliance is sequentially incentive compatible if and only if immediate slashing plus discounted reputation loss is at least $\Gamma$ at every reputation tier. We use this condition to formulate a protocol design problem mapping service primitives to stake requirements, reward schedules, probation rules, and audit frequency.
Read the full abstract
Data Structures and Algorithms
Arithmetic Variable LogLog: Advancing the Memory-Variance Frontier
Cardinality estimation - counting the number of distinct elements in a data stream - requires a tradeoff between memory and accuracy. ExaLogLog recently established the state of the art for this tradeoff by combining wide registers with a Fisher-information-optimal maximum likelihood (ML) estimator, achieving the best known memory-variance product (MVP) among HyperLogLog variants. Here we present Arithmetic Variable LogLog (AVLL), which surpasses ExaLogLog at every memory point tested using arithmetic encoding and eliminating uncommon states to consume 64-bit words completely with 11 registers each, yielding a 5.5x register-count advantage. Its four-component blended estimator, HLDLC, exploits this density advantage to surpass ExaLogLog's ML accuracy without iterative solving. At 1 KB, AVLL achieves 1.63% width-weighted mean absolute error compared to ExaLogLog's 1.71% - a 4.7% improvement. The corresponding empirical MVP is 3.4, surpassing ExaLogLog's practical MVP of 3.78 and its theoretical optimum of 3.67. This holds at every tested size from 0.25 to 4 KB. AVLL inherits DynamicLogLog's early exit mechanism, which filters most elements before any register is touched. With thousands of simultaneous sketches per thread, AVLL is 2.7-4.5x faster than ExaLogLog due to the reduced memory bandwidth from early exits. Like DynamicLogLog, AVLL stores relative NLZ values with a shared offset, so its memory scales as O(B + log log C) rather than O(B x log log C) - decoupling maximum representable cardinality from register width. These results hold under both high-complexity (all-unique) and low-complexity (nonuniformly high duplication rate) data distributions, with zero accuracy degradation from duplication. AVLL is implemented as a single self-contained Java class with all correction formulas embedded, available in the BBTools suite at https://bbmap.org.
Read the full abstract
Artificial Intelligence
Trie Automata for Constrained Decoding over Large Finite Sets
Large language models increasingly need to generate structured outputs that conform to predefined schemas, with one common constraint being selection from a finite set of valid strings. Current constrained decoding systems handle this through general-purpose grammar compilation, which becomes prohibitively slow as the number of valid values grows into the thousands, a cardinality wall. We introduce the trie automaton, a specialized mechanism that exploits finite-set structure (shared prefixes, bounded depth, known cardinality) via Aho-Corasick multi-pattern matching to precompute per-node token masks. The trie achieves 7X faster per-step valid-token computation (0.65 us vs. 5.8 us) compared to XGrammar, one of the primary backends in vLLM and SGLang, and 2–6.5X faster compilation at K >= 300. Because precomputed masks enable a stateless serving path that bypasses the guided decoding pipeline, this advantage compounds in batch serving: end-to-end vLLM throughput reaches 219 req/s vs. XGrammar's 7.5 req/s at batch size 256 (29X). The 29X combines the algorithmic speedup with integration-path savings that only precomputed masks can unlock. Across seven tokenizer families (32K–262K vocabulary), the trie maintains sub-100ms compilation up to K = 10,000 and flat per-step cost regardless of set size, while guaranteeing 100% output validity.
Read the full abstract
Machine Learning
Prof-K: Probabilistic One-Pass Filtering for Efficient Top-k Selection
Top-k selection is a fundamental computational primitive with applications spanning databases, information retrieval, signal processing, and modern machine learning workloads, including sparse activations and attention pruning. As data sizes grow, existing approaches become inefficient: exact methods incur high memory and compute overhead, while approximate methods often rely on brittle heuristics that degrade under adversarial or heavy-tailed inputs. In this paper, we introduce Prof-K, a fast, scalable, and distribution-agnostic algorithm for exact top-k selection. Prof-K performs a single-pass filtering procedure: a small random sample estimates an adaptive threshold, the N input elements are streamed once into a compact buffer, and an exact top-k routine on this buffer recovers the true top-k elements on the first attempt with probability at least $1-\varepsilon$, where $\varepsilon>0$ is user specified. We derive high-probability guarantees for correctness and buffer size, together with an approximately optimal sample size that minimizes overhead as a function of N and k. Empirically, Prof-K achieves 1.5x-15x speedups over the highly optimized PyTorch topk and recent RadiK implementations, with the largest gains in the large-scale, small-to-moderate-k regime where prior methods struggle most. Unlike previous approaches, these guarantees hold independently of the input distribution, ensuring robustness to adversarial settings. A run-time check detects the rare failures and triggers a retry, so the returned set is always exact and $\varepsilon$ bounds only the probability of requiring an additional pass. We further demonstrate its impact on training BatchTopK Sparse Autoencoders (SAEs), where top-k selection constitutes a significant portion of the training cost.
Read the full abstract
Digital Libraries
Is this Citation on Point?
In 2023, a New York judge sanctioned two attorneys in Mata v. Avianca for filing a brief containing hallucinated citations generated by ChatGPT. These failures can be largely caught by database lookups. The harder problem is detecting citations that point to real cases but do not support the propositions for which they are offered, a failure mode that existing evaluations of LLMs for legal use cases largely overlook. In this paper, we study proposition-level citation support verification through controlled perturbations of real legal citations obtained from two legal corpora. We either replace the cited case or change only the pinpoint page within the same case. Our evaluation of fourteen model configurations on the resulting examples shows that models catch $93$–$100\%$ of wrong-case corruptions, but they only catch $37$–$61\%$ of wrong-pinpoint corruptions on court opinions and $52$–$83\%$ on legal briefs. When models fail to catch wrong-pinpoint corruptions, they accept the citation based on topical overlap rather than page-level support. Larger-scale models and extended reasoning narrow the gap but do not close it: GPT-5.4 with high reasoning effort still misses $40\%$ of pinpoint mismatches on court opinions and $18\%$ on briefs. Prompting the model to verify support at the cited page improves recall, but it also raises the false positive rate. These results show that citation misrepresentations remain difficult to detect automatically. Recognizing the right legal topic and verifying support for the cited proposition are distinct capabilities, and current models conflate them.
Read the full abstract
Computer Vision
Attribute-Conditioned Multimodal Slot Factorization for Controllable Fashion Retrieval
Fashion retrieval often requires satisfying multiple attributes at once, such as category, color, pattern, and demographic. Monolithic embeddings mix these signals into a single vector, making attribute-specific control difficult at retrieval time. Many existing semantic-ID methods provide discrete item codes, but these codes are typically optimized as item-level or residual addresses and do not expose named, independently controllable attribute slots. We introduce MM-slotgate, a multimodal slot encoder that factorizes Fashion-CLIP text and image embeddings into four named attribute slots. Each slot learns its own text-image gate, so visually grounded attributes such as color and pattern can rely more on image evidence, while taxonomy-oriented attributes such as category and demographic can remain more text-driven. On H&M, using a combined slot-similarity and slot-logit retrieval score, MM-slotgate achieves 0.7566 macro ConstraintSatisfied@10, outperforming equal-weight multimodal fusion (0.7142) and fCLIP text-only retrieval (0.4755). The largest gain is on color, which improves from 0.321 to 0.889 (+0.568 absolute), as the learned color gate assigns 57.4% weight to image evidence. The learned gates are interpretable without modality supervision: color is image-leaning, category is text-leaning, and pattern and demographic lie near the middle. The resulting slots also remain controllable: linear probes show no measured excess leakage beyond the label-correlation baseline, and quantized slot codes support targeted intervention, including a 15.3x lift for color. These results suggest that controllable fashion retrieval benefits from typed, attribute-conditioned multimodal slots rather than either a single global embedding or opaque item-level semantic IDs.
Read the full abstract
Information Retrieval
Test-Time Optimization of Query Embeddings with Ranking Aware Reward Maximization
Dense retrievers rank documents using vector similarity between a frozen encoder and a precomputed index. While test-time ranking rewards from a reranker or LLM judge can improve results, existing methods discard this signal after a single query. Updating the retriever's weights makes rewards reusable, but this requires parameter access, which is unavailable for closed-source models, and is computationally prohibitive. We propose TTT-Embed (Test-Time Tuning of Embeddings), a framework that distills ranking rewards into a lightweight, learned vector within the output embedding space of a frozen model. This vector is optimized purely from scalar ranking scores assigned to the retriever's own candidate documents, requiring no access to model weights, ground-truth labels, or modifications to index. A single scope parameter controls rewards reuse (global, task, or query), enabling a principled trade-off between reusability and specificity under a fixed reward computation budget. We demonstrate that as the available reward budget scales, the optimal sharing scope shifts dynamically from global-wise to task-wise and finally to query-wise. Evaluated across five embedding models and 15 MTEB retrieval tasks, TTT-Embed improves test-time retrieval by up to +8.36 nDCG@10. Crucially, the learned states generalize effectively to unseen queries (up to +8.57 nDCG@10) and unseen tasks (up to +4.71 nDCG@10). Furthermore, TTT-Embed successfully resolves catastrophic forgetting: by leaving base weights entirely frozen, it recovers degraded general capabilities (up to +8.00 nDCG@10, even surpassing the original base model) while preserving in-domain specialization. These results establish ranking rewards as a reusable test-time state, enabling budget-efficient adaptation for any embedding model, including closed-source APIs.
Read the full abstract
Emerging Technologies · Cross-listed from cond-mat.mes-hall
Spin Vector Control for Heisenberg-Inspired Probabilistic Computing
Probabilistic bits (p-bits) have emerged as a cornerstone of probabilistic computing, enabling energy-efficient hardware implementation for probabilistic inference and combinatorial optimization. A critical challenge in advancing this field beyond binary p-bits lies in realizing and manipulating vector spin information, essential for mapping complex energy-based models such as the Heisenberg Hamiltonian.Here, we demonstrate a spintronic platform capable of real-space vector summation by using dual ferromagnetic spin injections into a monolayer graphene channel. By electrically tuning the spin polarization through independently controlled injection currents, we achieve continuous control over the magnitude and direction of the resulting spin accumulation vector. Experimental observations, supported by theoretical vector summation models and spin-circuit simulations, reveal coherent vector interactions and angular tunability of the spin state. This approach enables direct implementation of vector-based spin logic and lays the groundwork for mapping classical Heisenberg models using stochastic low-barrier magnets. Our results establish a scalable pathway for realizing probabilistic spin circuits based on two-dimensional materials, offering new opportunities for low-power, non-Boolean computing architectures.
Read the full abstract
Digital Libraries
Domus: An Open-Data Web Platform for House History Research
House history – the record of who lived in a building, who owned it, and how its address changed over time – is a central concern of genealogical and local history research. Wikidata and OpenHistoricalMap together provide an open-data infrastructure, CC0-licensed and capable of representing this data: Wikidata through its linked-data property model with temporal qualifiers, OpenHistoricalMap through historical building footprints with start and end dates. What has been missing is a domain-specific tool that makes contributing and discovering building history accessible to genealogists and local historians. Domus fills this role: a map-based web platform that guides users through structured, crowdsourced editing of building records in Wikidata and displays corresponding OpenHistoricalMap footprints. The application requires no custom backend – authentication, data reading, and data writing all operate through public APIs in the browser, and all data are stored under CC0 in public collaborative databases. This paper describes the genealogical data model, the architecture, and key technical design decisions.
Read the full abstract
Machine Learning
Scaling Automatic Research Agents via World Models
Automating empirical research is a long-standing direction of AI. Recent automatic research (AutoResearch) agents bring this goal within reach, as modern LLMs show the capability to independently implement solutions and learn from the execution outcomes. Behind these gains, post-training (especially RL) plays a central role. In this paper, we identify a fundamental tension when scaling RL for these agents: the two components of every AutoResearch trajectory (agent generation and environment execution) scale in very different manners, since all generation shares compute through batching, while each execution occupies its exclusive sandbox and real machine time. As a result, the environment execution dominates the training cost and becomes the bottleneck as trajectories grow. To resolve this tension, we propose World Model RL (WMRL), which replaces environment execution with a world model to remove this bottleneck. Additionally, the world model can be imperfect, as its rewards are corrupted by bias and noise. Therefore, we further equip WMRL with two mitigations, Online Debiasing and Inverse-Variance Denoising, which offset the bias and suppress the noise respectively. Theoretically, we prove that both mitigations of WMRL strictly improve the convergence guarantee. Empirically, WMRL accelerates training by 3-4x on various tasks at different agent scales, while exceeding the performance of standard RL baselines. Moreover, our post-trained 4B and 9B agents outperform much larger open-weight agents of 48B and 120B on held-out benchmarks. Beyond AutoResearch, WMRL also transfers to post-training embodied VLA policies, which demonstrates the generalizability of our method.
Read the full abstract
Artificial Intelligence
CAS: A Causal Attribution Score for Local and Global Explainable Artificial Intelligence
Predictive explanation methods attribute a model output; they do not, by themselves, attribute an intervention effect on the real-world outcome. We introduce the Causal Attribution Score (CAS), a compact score architecture for causal explanation. CAS starts from an identified interventional coalition game, allocates the joint intervention contrast with causal Shapley contributions, and converts those raw outcome-scale effects into Local CAS, Signed Local CAS, and two complementary Global CAS summaries. The innovation is not a new Shapley formula, but a local-to-global causal reporting layer with an explicit intervention target. In the known-truth benchmark, eight repeated primary-interaction simulations (n = 2,200 each, three actions) gave mean Local CAS MAE of 0.107 for coalition-aware CAS, compared with 0.173 for one-at-a-time normalisation and 0.213 for a global normalised absolute ATE vector. The paired advantage over one-at-a-time normalisation increased from -0.003 under additivity to 0.091 under strong interactions. On both empirical DoubleML datasets, 401(k) eligibility/net financial assets (n = 9,915) and Pennsylvania reemployment bonus/unemployment duration (n = 5,099), predictive SHAP/TreeSHAP rankings differed materially from Feature-CAS rankings of treatment-effect modifiers. In Pennsylvania, dep1 (exactly one dependent) moved from predictive global rank 13 to Feature-CAS rank 2 and was the leading local Feature-CAS modifier. These results isolate the added value of separating what predicts the outcome from what explains heterogeneity in an estimated causal effect.
Read the full abstract
Human-Computer Interaction
Legally Mandated, but Still Inaccessible: Digital Tensions in Older Adults' Use of Norwegian Web Services
Norway is among the most digitalized countries in the world, where access to essential services increasingly depends on digital systems. Although universal design of ICT is legally required across public and private sectors, ensuring cognitive accessibility for older adults involves more than technical compliance. We analyzed responses from 294 participants aged 55 to 90 to examine the barriers they encounter when using digital services. Our findings identify four tensions, namely navigational, semantic, procedural, and temporal, which reveal how accessibility barriers emerge through misalignments between service design, system assumptions, and older adults' capabilities. Together, these tensions show that checklist-based compliance does not necessarily translate into lived accessibility when engaging with mandatory or near-mandatory digital services. We discuss the need to move beyond minimum compliance frameworks toward accessibility approaches that better support older adults' independent participation in digital society.
Read the full abstract
Computational Complexity
On the Exponential Circuit Imbalance of the Ben-Tal Nemirovski Approximation
Dadush et al.\ (2024) recently developed a scaling-invariant layered least squares algorithm for linear programming whose complexity depends on the optimal condition measure $\bar{\chi}_A^*$. Their work builds on Vavasis and Ye's (1996) algorithm whose running time depends only on the constraint matrix $A$ through the condition number $\bar{\chi}_A$. Monteiro-Tsuchiya (2003) defined the optimal condition number $\bar{\chi}_A^*$ as the maximum $\bar{\chi}_{AD}$ achievable over all positive diagonal column rescalings $D$. Dadush et al.\ (2024) introduced the optimal circuit imbalance measure $\kappa_W^*$, which serves as a lower bound for $\bar{\chi}^*_A$. Instances with artificially large optimal circuit imbalance measures $\kappa_W^*$ can be easily constructed; however, finding naturally occurring examples where this optimal scaling-invariant measure grows exponentially is of independent interest. In this paper, we show that the Ben-Tal Nemirovski (BN) linear programming approximation of the unit disk provides such an example. By explicitly constructing circuits in the kernel of the BN formulation, we prove that the optimal circuit imbalance measure $\kappa_W^*$ grows exponentially in the number of approximation steps. Since $\kappa_W^*$ lower bounds $\bar{\chi}_A^*$, our result demonstrates that the BN approximation yields an exponentially ill-conditioned family of constraint matrices.
Read the full abstract
Computer Vision
StrAD: A Streaming Method and Benchmark for Audio Description Generation for Long-form Videos
Visual content is the dominant medium of communication, yet without audio descriptions (ADs), it remains inaccessible to blind and low-vision people. ADs narrate context-relevant visual events during natural audio pauses. Manually creating ADs is expensive, limiting coverage to a small fraction of available content. Most existing automatic AD generation methods frame the task as video clip captioning, requiring ground-truth timestamps and additional context cues such as character databases. Current benchmarks reinforce this framing, consisting of short video segments paired with automatic or task-mismatched annotations. We introduce StrAD, a benchmark for long-form AD generation on full-length videos spanning diverse genres such as movies, documentaries, short films, performances, and video games. We reformulate AD generation as streaming dense video captioning. Our approach processes full-length videos with a sliding window, inserting ADs into existing transcripts without ground-truth timestamps, and supports both fine-tuned models and zero-shot prompting of vision-language models. On the segment-level task with given timestamps, our fine-tuned StrAD-FT sets the state of the art on CMD-AD with 36.3 CIDEr (+10.0 over Shot-by-shot), establishes a reference point on StrAD (51.0 CIDEr), and remains competitive on MAD-Eval at 24.9 CIDEr. On the full-video streaming task, StrAD-FT reaches a SODA score of 2.4 against 1.1 for our zero-shot baseline StrAD-Zero, though both exhibit limitations in temporal localization and narrative coherence. While prior work has tackled full-video AD generation in an offline, multi-stage fashion, ours is the first streaming approach, generating ADs on the fly without ground-truth timestamps. StrAD makes progress on full-video AD generation measurable, a prerequisite for scaling accessibility.
Read the full abstract
Human-Computer Interaction
Analysis of Motor Signatures of Social Adaptation in Autism for Efficient Human-Centric Systems
Dance imitation integrates motor planning, sensorimotor integration, and social cognition, offering a sensitive framework to characterize motor behavior in autism. In this work, we explore a computational analysis framework to identify potential biomarkers that allow the design and development of improved medical and human-machine systems. We analyzed 3D motion capture data from autistic and neurotypical adults performing dance imitation under solo and socially-framed duo conditions. Methodologically, using Dynamic Time Warping, we quantified movement consistency and propose the Social Context Sensitivity Index (SCSI) to measure modulation of variability by social framing. These features were then used on a classifier to discriminate subjects into autistic or neurotypical groups. Results show that neurotypical adults exhibited increased movement variability in socially-framed imitation, especially in upper and lower limbs, whereas autistic adults maintained consistent movement across contexts. Classification achieved 79.2% balanced accuracy in distinguishing groups. These findings suggest that social context sensitivity in motor imitation constitutes a robust biomarker of autism-related motor behavior, highlighting the importance of social modulation in motor assessments and informing the development of inclusive human-centric technologies.
Read the full abstract
Multiagent Systems
Do LLMs Beat Nash? Testing Decentralized Coordination in Self-Play Multi-Agent Games
Large language model agents deployed without a central controller are often assumed to require communication to coordinate their actions. We ask what remains possible without it: when independent instances of the same model cannot communicate, can they still reason about their counterparts well enough to exceed the standard game-theoretic baseline for uncoordinated play? We introduce a benchmark of one-shot, no-communication games in which each of thirteen language models is told only that its counterparts are running the same model and is evaluated against the Nash equilibrium of the underlying game. In two-player matrix games spanning seven archetypes and two to ten actions per player, two frontier-hosted models consistently exceed their Nash benchmark, approaching the optimal joint outcome in several archetypes, while most open-weight models achieve only partial gains that vary sharply by game structure. Performance degrades substantially in team-based games with four or more interchangeable agents, particularly as the action space grows, suggesting that whatever capability drives self-play gains in dyadic games does not transfer to larger multi-agent teams.
Read the full abstract
Human-Computer Interaction
Transforming Interactions in Thesis Supervision: An Exposé-First Workflow in Higher Education
At the studied research institute, one professorship oversees approximately 20 theses per semester, while day-to-day supervision is distributed among doctoral and postdoctoral researchers. To manage this supervision demand, the institute uses an exposé-first workflow in which students prepare a research proposal before entering the main thesis-writing phase. This paper asks how students, supervisors, and administrators experience the exposé-first workflow as a structured process for early thesis preparation, and how it redistributes responsibility, supervision, and administrative coordination work across roles and two digital platforms. Based on a mixed-methods study analyzed through Frauenberger et al.'s four reflective design lenses, the findings show that the exposé-first model made thesis preparation more structured by turning early research planning into a staged process of proposal writing, feedback, and approval. Students reported that this process helped them clarify research goals and take ownership of their research plans at an early stage. However, the workflow redistributed rather than reduced work: supervisors shifted toward iterative feedback, feasibility checking, and preliminary quality assurance, while administrators carried much of the coordination across platforms, deadlines, submissions, and feedback. The paper contributes an analysis of exposé-first thesis preparation as a sociotechnical workflow, showing how workflow redesign can improve structure while leaving essential administrative coordination work underrecognized.
Read the full abstract
Computer Vision
Surface-to-Skeleton 3D Cephalometry: Estimating Hidden Skeletal Landmarks from CT-Derived External Soft-Tissue Surfaces
Existing 3D facial-landmark methods localize points on visible skin, but whether CT-defined internal skeletal landmarks can be inferred from external soft-tissue geometry remains unclear. We formulate a coordinate-consistent surface-to-skeleton task using same-acquisition CT-derived surfaces, separating estimation from optical-to-CT registration, scanner-domain, and acquisition-state effects, with coverage analyzed separately. From 240 clinical CT scans from two hospitals, we construct a locked retrospective protocol pairing CT-derived external soft-tissue point clouds with 21 skeletal landmarks and three visible soft-tissue landmarks. An integrated hierarchical point-cloud model achieves 2.97 mm mean radial error on skeletal landmarks and 3.03 mm on deep or surface-invisible landmarks in 40 held-out patients. Patient-mismatch controls support patient-specific signal beyond a fixed population configuration or global similarity alone, while coverage ablations indicate dependence on non-anterior geometry. Optical-transfer diagnostics reveal substantial coverage-related and global-configuration components, although deployable optical inference remains unresolved. These results answer the controlled feasibility question affirmatively and provide a basis for hidden skeletal landmark inference.
Read the full abstract
Machine Learning
GENADA: efficient generative time series adversarial attack framework
Deep learning models are widely used for time series analysis in domains such as healthcare, finance, energy systems, and environmental monitoring. However, these models remain vulnerable to adversarial attacks, where small input perturbations cause severe degradation in predictive performance. Commonly used gradient-based attacks, iterative first-order methods, are computationally burdensome, as they repeatedly backpropagate through the victim model to compute input gradients during a number of iterative refinement steps. We propose a GENerative ADversarial Attack (GENADA) that learns a generative model to produce deceptive perturbations directly in a single forward pass and a procedure to train it. Variants include single-step and iterative generative attack schemes. The validation considers attacks on several neural models and datasets in the time-series domain, a controlled, low-dimensional setting. Empirically, GENADA achieves comparable attack quality to strong baselines while requiring less time to generate perturbations during inference.
Read the full abstract
Multiagent Systems
Entropy-Augmented Multi-Objective Policy Optimization in Multiagent Systems
Autonomous agent teams deployed in settings such as marine and extraterrestrial outposts must coordinate actions to achieve optimal outcomes across multiple competing objectives. Multi-objective evolutionary algorithms such as NSGA-II optimize for diversity in the objective space, but neglect diversity in the behavior space, possibly leading to premature convergence and a collapse in behaviors that may differentiate policies in different external conditions. To address this, we introduce an entropy-augmented policy evaluation strategy that incorporates an entropy bonus into agent fitness scores, discouraging behavioral homogeneity across the evolving population. By augmenting policy evaluation with a behavior-space diversity signal while preserving the underlying Pareto optimization framework, our method is designed to encourage exploration of behaviorally distinct policies in multiagent domains. We evaluate our approach across rover-domain experiments with qualitatively distinct reward structures and observe hypervolume improvements of up to 48% relative to the NSGA-II baseline, suggesting that behavioral diversity is a promising and underexplored direction for improving multi-objective multiagent evolutionary optimization.
Read the full abstract
Multimedia
MASCOT: Model-Aware Submodular Coverage for Composite-Attribute Text-to-Image Retrieval
Vision-Language Models (VLMs) are highly effective in retrieving semantically relevant images. However, in practice, relevance alone is often insufficient. Systems must also achieve Result Diversification (RD) across composite attributes such as geography and time, a task for which precise control remains challenging. Current re-ranking methods, such as Multi-Source Determinantal Point Processes (MS-DPP), address this using manifold-based repulsion over similarity representations. Although this strategy is effective for broad exploration, it exposes a key limitation in manifold-based models: when subjected to diversity-decrease tasks on discrete metadata, they suffer substantial degradation in early-rank recall. To bridge this gap, we introduce MASCOT (Model-Aware Submodular Coverage for Composite-Attribute Text-to-Image Retrieval). Instead of relying on manifold repulsion, MASCOT formulates multi-attribute diversity as a resource allocation problem, projecting attributes into a soft-binning space weighted by query-driven importance. Averaged across the three PixelProse diversity-decrease tasks, MASCOT preserves an early-rank recall (R@10) of 88.58%, while MS-DPP retains 67.63%. The margin widens under composite constraints: on PP_geo_hour, where temporal and geographic diversity must be suppressed simultaneously, MS-DPP's recall collapses from 0.9737 to 0.4931 and its top-ranked result degrades to R@1 = 0.23, while MASCOT holds R@10 = 0.9410 and R@1 = 0.7202 at a diversity metric above the unconstrained baseline. We do not claim uniform superiority: on aggregate diversity-relevance scores our own simpler ablations attain higher harmonic means on all three decrease tasks, and MASCOT's advantage is specific to recall beyond rank 1 under composite constraints.