Artificial Intelligence · Cross-listed from cs.CY
Prepared Or Unprepared? Evaluating Healthcare Workforce Readiness for Clinical Adoption of Artificial Intelligence in Nigeria
Abbas M. Rabiu, Abdulrazaq A. Zubair, Um-mulkhairi Ibrahim et al. (+6 more)
Artificial intelligence (AI) is increasingly integrated into healthcare systems worldwide, yet its successful clinical adoption depends critically on workforce readiness, particularly in low- and middle-income countries (LMICs) where infrastructural and training gaps persist. This cross-sectional study evaluated awareness, attitudes, preparedness, and barriers to AI adoption among 761 healthcare professionals across multiple disciplines and practice settings in Nigeria. Data were collected between December 2025 and March 2026 using a structured, validated questionnaire. Overall awareness of AI in healthcare was high (92.6%); however, objective knowledge and self-reported preparedness remained limited, with 40.9% reporting low or very low knowledge and only 63.0% feeling adequately prepared. Willingness to adopt AI was high: 92.5% expressed interest in training, and 78.7% supported inclusion of AI education in undergraduate curricula. Key barriers included lack of training (84.7%), poor infrastructure (71.1%), high cost of AI tools (61.0%), fear of job displacement (60.6%), ethical concerns (52.9%), and data
privacy concerns (52.7%). Significant differences in preparedness were observed across geopolitical zones (chi-square (5) = 24.28, p < 0.001), and awareness differed across professional groups (chi-square (6) = 68.38, p < 0.001). Attitudes toward AI differed significantly across professional groups (F = 3.32, p = 0.003), with professionals who felt prepared demonstrating more positive attitudes (mean = 3.74) compared to those who did not (mean = 3.46). These findings reveal a critical disconnect between high awareness and actual readiness, underscoring the need for targeted training, infrastructure investment, and clear implementation frameworks to bridge the gap between AI technological potential and clinical reality in resource-constrained settings.
Read the full abstract ▾
All authors Abbas M. Rabiu, Abdulrazaq A. Zubair, Um-mulkhairi Ibrahim, Tolulope Olusuyi, Shaheeda Farouq, Safwan M. Dafi, Adaobi C. Emegoakor, Yewande Gbadamosi, Maruf Adewole
Artificial Intelligence · Cross-listed from cs.CL
Reporting Practice Matters: The Impact of Reference Choice on Chest X-ray Report Evaluation
Daniel P. Jeong, Charles Q. Li, Hossein Hosseiny et al. (+5 more)
Radiologists follow heterogeneous reporting practices. Two radiologists examining the same image and identifying the same clinical findings might nevertheless compose superficially distinct reports, varying in terminology, shorthand, formatting, and level of detail. These variations in reporting norms represent an under-appreciated obstacle in efforts to evaluate AI-based radiology report generation (RRG) models, where machine-generated reports are typically assessed based on their concordance with human-generated references. In this paper, we quantify the sensitivity of established evaluation metrics to variations in reporting practices, revealing impacts large enough to alter the rankings of models. We introduce a radiologist-informed taxonomy of variations in radiology reporting practice and a method (ReRef) that rewrites reference reports along the axes of our taxonomy while preserving clinical interpretation. For instance, when comparing the performance of nine RRG models on MIMIC-CXR using RadCliQ-v1, condensing the discussion of normal findings in the reference reports causes Libra to drop from first to second place while CheXOne rises from third to first. Our results suggest that many current metrics fail to decouple clinical interpretation from conformity to reporting practices and that choosing the “right” references that accurately reflect the desired reporting practices can be important in practice. To support future research, we release MIMIC-CXR-Ext-ReRef, a radiologist-validated
dataset of 120 (original, alternative) reference report pairs derived from MIMIC-CXR.
Read the full abstract ▾
All authors Daniel P. Jeong, Charles Q. Li, Hossein Hosseiny, Nitya M. Bhalla, Fatma Uyar Morency, Pradeep Ravikumar, Zachary C. Lipton, Michael Oberst
Artificial Intelligence · Cross-listed from quant-ph
Securing quantum error correction against misleading advice from AI agents
A. Bar{\i}ş Özgüler
Can an attacker turn influence over an artificial intelligence (AI) adviser into a harmful
quantum error-correction update? We identify an ambiguity in passive syndrome records that obstructs recovery selection, then show how additional calibration measurements support certified recovery updates under uncertainty and drift. In an odd-distance square toric code with error-free preparation, syndrome measurements, and recovery operations, opposite coherent $X$ rotations produce identical passive syndrome-history distributions. Yet a fixed phase correction can help at one sign and harm at the other. A terminal logical measurement on known encoded calibration states supplies the missing sign information. A separate evaluator accepts an update only when calibration uncertainty and a justified drift bound certify improvement over the current recovery, without assuming that the adviser recommends correctly. In simulated advice attacks, calibration-confidence checks reject harmful proposals while retaining beneficial updates under honest advice. We derive sufficient limits on calibration age that require improvement through deployment. In matched simulations, a validated channel-specific bound retains more beneficial updates than the general bound after accounting for evaluation time, while preventing the tested harmful activations under the stated drift assumption. A separate surface-code experiment includes
stochastic circuit faults and noise changing during acquisition. Deterministic controllers achieve at least as many beneficial updates with the same observations. Violating the drift assumption permits harmful acceptance in the toric experiment. The results identify information required for recovery selection, establish conditional guarantees against harmful updates, and quantify the recovery improvements forgone through conservative acceptance.
Read the full abstract ▾
Artificial Intelligence
MUSE: Benchmarking Large Vision-Language Models on Multi-Modal Understanding in Situated Education
Luyao Zhu, Xun Wei Yee, Wei Li et al. (+2 more)
Large
vision-
language models have achieved remarkable progress in multi-modal understanding, yet their capabilities in educational settings remain insufficiently evaluated. In AI-assisted
language learning, models must interpret artistic imagery, understand its semantic, affective, and cultural content, and reason about visual context to support meaningful interaction. However, existing
benchmarks primarily focus on real-world images or domain-specific educational
reasoning, providing limited coverage of artistic educational content. To address this gap, we introduce MUSE, a
benchmark for evaluating large
vision-
language models on artistic image understanding in situated educational applications. MUSE decouples image annotation from question generation, enabling diverse tasks with controllable difficulty while reducing annotation effort. It comprises twelve tasks spanning visual perception, semantic and affective interpretation, culture understanding, and compositional
reasoning, together with diverse artistic images deliberately curated to center Singaporean and Southeast Asian multicultural contexts alongside Western art traditions, covering multiple themes and difficulty levels. Evaluation of open-source and proprietary models reveals substantial disparities across capability dimensions, particularly in affective interpretation and compositional
reasoning. Our analysis further identifies common failure modes and key challenges for developing trustworthy multi-modal models for education. We hope MUSE will serve as a standardized
benchmark for advancing multi-modal understanding in situated educational applications.
Read the full abstract ▾
All authors Luyao Zhu, Xun Wei Yee, Wei Li, Mun Thye Mak, Wee Siong Ng
Artificial Intelligence · Cross-listed from cs.LG
Probabilistic Linear Explanations
Frederic Koriche, Jean-Marie Lagniez and Chi Tran
Formal explainability provides mathematically grounded justifications for individual predictions. However, abductive explanations often exceed human cognitive limits by involving too many features, while probabilistic relaxations have remained largely limited to categorical classification. We present a unified framework for probabilistic explainability based on sparse, anchored linear models, applicable to both binary classification and continuous regression. By mapping instances to the Boolean hypercube, our linear explanations strictly generalize subset-based approaches: they capture both the magnitude and direction of feature contributions while enforcing a prescribed sparsity budget $k$. We show that minimizing the relevance error for such explanations is \ClassNPPP-hard when the underlying model is a neural network, and we relate this intractable objective to a tractable surrogate—the fidelity error. For a parameterized family of local distributions, the relevance error of any $k$-sparse explanation is bounded by its fidelity error up to a multiplicative factor that remains small locally. We address the resulting empirical problem using two complementary approaches: a Mixed Integer Programming (MIP) formulation that yields provably optimal empirical solutions while maintaining
polynomial sample complexity, and a
polynomial-time Iterative Hard Thresholding (IHT)
algorithm with provable approximation guarantees. Empirical evaluations show that, unlike state-of-the-art baselines such as LIME and MAPLE, our explanations satisfy both the anchoring and sparsity constraints by construction, while consistently achieving lower relevance error.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.LG
Double descent is the principle of least action
Congzhou M Sha
The test error of a model plotted against its number of parameters $d$ falls, peaks when the model can just fit the training data, and falls again, exhibiting the double descent phenomenon. We explain the phenomenon with statistical mechanics. The training trajectory of a
stochastic gradient-based method is a particle wandering over the energy landscape of the training loss at an induced temperature $T$, and a run that has equilibrated visits every parameter vector of a given training loss equally often, the fundamental postulate of statistical mechanics, with probability given by the Boltzmann distribution. Because training starts at an initial point and has only finite time to diffuse, it carries an effective weight decay, which makes every parameter a quadratic degree of freedom. The equipartition
theorem then distributes the energy among the $d$ degrees of freedom in shares of $T/2$, so at a fixed training loss adding parameters lowers the temperature and drives the Boltzmann distribution toward the stationary path. Finally, adding parameters can only lower the $L^2$ norm of the stationary path, so a solution sampled at fixed loss is less likely to be large with increasing $d$, effectively increasing weight regularization.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.LG
RLLBC-Lib: An Educational Code Library for Reinforcement Learning and Learning-Based Control
Bernd Frauenknecht, Emma Cramer, Artur Eisele et al. (+9 more)
Reinforcement learning (RL) is an exciting concept as well as a remarkable success story worth sharing. However, RL builds on rather complex interactions between different objects that play out over several cycles. Such dynamics are often best explained with an easily accessible implementation. We present RLLBC-Lib, a carefully crafted code library with the goal of lowering the entry barrier for students and other learners of RL in the context of learning-based control. At its heart, RLLBC-Lib comprises a comprehensive library of tabular RL approaches to enforce a clear understanding of the theoretical foundations. A deep RL library follows the same design principles, underscoring the parallels between simple tabular and state-of-the-art deep RL approaches. Additionally, RLLBC-Lib provides a collection of implementations illustrating core RL principles and contrasting RL to other learning-based control approaches. Finally, RLLBC-Lib provides an ideal basis for creating programming assignments with automated grading.
Read the full abstract ▾
All authors Bernd Frauenknecht, Emma Cramer, Artur Eisele, Paul Kruse, Lukas Kesper, Jonas Hertrampf, Ramil Sabirov, Jyotirmaya Patra, Johannes Berger, Paul Brunzema, Friedrich Solowjow, Sebastian Trimpe
Artificial Intelligence · Cross-listed from cs.MA
Social Laws for Multi-agent Coordination in Stochastic Environments
Rolando Fernandez, Caleb Probine, Tyler Lee et al. (+5 more)
In multi-
agent environments, coordinating
agents to prevent interference and ensure robust individual performance is a critical challenge. Previous research on social laws for multi-
agent systems has primarily focused on deterministic, goal-based settings. This paper extends the concept of social laws to
stochastic, reward-based environments, proposing a formalism for defining and verifying their robustness under various conditions. We introduce the notion of $\alpha$-robustness, a measure of the guaranteed utility each
agent retains while pursuing its optimal single
agent policy, assuming all
agents obey the social law. We then present an approach for robustness verification of social laws in
stochastic settings, based on a reduction to solving a series of Markov decision processes. Empirical evaluations on toy environments illustrate the potential of our framework.
Read the full abstract ▾
All authors Rolando Fernandez, Caleb Probine, Tyler Lee, Jeffrey Chen, Erez Karpas, Muhammad Arrasy Rahman, Peter Stone, Ufuk Topcu
Artificial Intelligence · Cross-listed from cs.LG
Higher-order pruning of experts in mixture-of-experts language models
Alex M. Tseng, Prannay Kaul, Luca Zancato et al. (+2 more)
Mixture-of-Experts (MoE)
language models suffer from large parameter counts, which create a significant
memory bottleneck. Expert
pruning is the most direct approach for reducing this parameter count, yet existing methods make
pruning decisions for each expert independently, and assume experts' contributions are purely additive. In reality, expert usage in MoEs is inherently cooperative. We derive HOPE (Higher-Order
Pruning of Experts), a second-order
pruning objective which provably minimizes an upper bound on the error resulting from
pruning. We show that REAP (a state-of-the-art first-order
pruning method) is a special case of HOPE where interaction terms are ignored. Across three frontier MoE models (up to 122B parameters), two distinct calibration sets, and multiple
benchmarks (including math, instruction following, coding, and an agentic suite), we demonstrate that HOPE produces better
pruning decisions than existing methods, and its advantage is most pronounced at high
pruning rates and on challenging agentic workloads. At 50%
pruning, HOPE outperforms all baselines and achieves an average rank of 1.58 out of 5 methods (versus 2.42 for the next-best method, REAP), with gains of up to +6.1% on agentic coding. Over all conditions, HOPE again achieves the best average rank and surpasses every other method in the majority of head-to-head comparisons. By preserving cooperative expert structure that first-order methods ignore, HOPE enables aggressive compression with minimal degradation, particularly on complex tasks where diverse expert combinations are invoked over long sequences.
Read the full abstract ▾
All authors Alex M. Tseng, Prannay Kaul, Luca Zancato, Wei Xia, Stefano Soatto
Artificial Intelligence · Cross-listed from cs.CL
Beyond Outcomes: Dual-View Relational Learning for Efficient Agent Benchmarking
Xinshuai Guo, Junjie Wu, Dolly Deng et al. (+4 more)
Agent benchmarks are substantially more costly to evaluate than conventional LLM
benchmarks.
Benchmark compression is therefore a natural solution, yet existing methods primarily model redundancy in task–model final-score distributions, which is important in agentic evaluation. To address this limitation, we analyze large-scale trajectories and identify six complementary process signals that are systematically associated with final
agent performance. To disentangle
agent performance redundancy from a complete perspective, we propose DualViewEval, an
agent benchmark compression method that jointly exploits outcome and process relations to learn an exact-size miniset and predict the full-
benchmark scores. Across five
agent benchmarks and five representative baselines, DualViewEval achieves the best results in all
datasets. With only 20 tasks, it achieves $24\times$–$40\times$ compression on APEX-
Agents and BFCL, reducing mean absolute error (MAE) by $14.5\%$–$28.2\%$ over the strongest competitors while improving Kendall's $\tau$ by up to $7.2\%$ relative to EssenceBench on SWE-bench Verified. The selected minisets further reveal capability differences among different
agents, providing compact and diagnostic feedback for efficient agentic model development.
Read the full abstract ▾
All authors Xinshuai Guo, Junjie Wu, Dolly Deng, Yinghui Li, Hai-Tao Zheng, Suncong Zheng, Maxm Pan
Artificial Intelligence · Cross-listed from cs.CR
ASLEval: Measuring Privacy Exposure Displacement in LLM Agent Sessions
Guosen Wu, Huizhen Huang, Guoxiong Long et al. (+2 more)
Privacy evaluations of tool-using LLM
agents often inspect a designated action, final response, or attacker report. These local proxies can miss unauthorized exposure elsewhere in a multi-step session and lack common ground truth across outlets, reports, and tool paths. We introduce
privacy exposure displacement, the mismatch between a local evaluation proxy and target-grounded session exposure, and ASLEval, an authorization-aware framework that pre-registers a hidden target set, measures all declared visible exits, and reserves internal traces for diagnosis. Across multiple enterprise-style environments and independently implemented runtimes, we observe three recurring patterns. An expected-outlet-only view misses 46.9% of exposure recovered by the visible-exit union; attacker self-reports combine omissions with high false discovery; and schema-aligned internal evidence usually precedes visible exposure at the request/probe level. Reducing model-visible returns changes this path but can eliminate normal-task success. Independent human review supports the adjudication pipeline while identifying harder console and candidate cases. These findings motivate
benchmarks that declare the complete visible boundary, ground claims in pre-specified targets and authorization, and report
privacy together with task utility.
Read the full abstract ▾
All authors Guosen Wu, Huizhen Huang, Guoxiong Long, Tao Huang, Chen Hou
Artificial Intelligence · Cross-listed from cs.CV
Decodable but Misrouted: Sparse Features Uncover a Readout Gap in Vision-Language Models for Harmful Meme Detection
Girish A. Koushik, Diptesh Kanojia and Helen Treharne
When large
vision-
language models misclassify harmful memes, the failure may reflect missing internal evidence or an inability to route represented evidence to their outputs. We distinguish these cases in Gemma-3 and Qwen3.5 using sparse autoencoders, role-conditioned probes, causal interventions, and recovery experiments across six harmful content
benchmarks, with additional Spanish and Hindi-English code-mixed evaluations. Sparse readouts outperform native prediction on all six primary binary tasks: Qwen averages $0.740$ versus $0.432$ for native macro-F1, residual reconstruction reaches $0.486$, and Gemma improves from $0.532$ to $0.714$. These gains measure how accessible the label is to a supervised readout; they do not show that the model's native generation already applies such a decision rule. Under the evaluated scales, Qwen silent-feature ablation is $24-63$ times more probe-sensitive, whereas routed-feature patching on literal yes/no tasks is $16-140$ times more output-sensitive. Native-only threshold calibration explains much, but not all of the gap: on five tasks with matched probe scores, it recovers $69.8$% of the raw native-to-probe difference, while direct routing adds $0.094$ mean macro-F1 beyond calibrated native scoring. Joint gold-label, probe-KL, and pairwise LoRA supervision improves dedicated FHM prediction, but a gold-only adapter performs better on the shared seven-task mean. A case study of Gemma-3-12B on the Facebook Hateful Memes
dataset finds a distributed rank-32 image-
prompt interaction, reaching $0.756$ versus $0.685$ native macro-F1. Robustness controls show that the signal is not explained solely by accompanying OCR and depends on paired visual evidence, and that it extends beyond English. In many of the errors we study, the evidence is represented but does not reach the answer; therefore, routing is a common bottleneck in harmful meme classification.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.NI
Taming the Agentic RAN: Stability-Guaranteed Arbitration of Autonomous AI Agents in O-RAN
Seyed Bagher Hashemi Natanzi and Bo Tang
The O-RAN control plane is becoming agentic: autonomous AI
agents, deployed as rApps by different vendors, independently close control loops over shared
radio resources. We demonstrate on a live O-RAN system that this independence is unsafe. Two
agents with individually correct objectives, one protecting a latency SLA and one maximizing utilization for energy efficiency, jointly drive recurring opposing excursions of the shared resource partition that neither produces alone. Existing conflict-mitigation mechanisms presume a statically known application population and cannot govern
agents whose behavior emerges at run time. We present AURA, a lightweight arbitration layer that admits
agent actions only when they satisfy feasibility invariants, per-variable dwell times, and a deadband, and we prove the arbitrated system converges to a feasible operating point. Implemented on an OpenAirInterface (OAI) testbed with measured one-way latency and throughput, AURA reduces recurring shared-state excursions by more than an order of magnitude (from 8.4 to 0.4 PRB amplitude) and virtually eliminates cross-slice throughput starvation (from 40-55% to 0.3%), while leaving the protected slice's own latency compliance unchanged, a trade-off the convergence guarantee makes explicit.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from eess.AS
GrainSpeech: Less Context, More Detail for Compact Speech Synthesis
Zitao Liang and Chang Gao
Compact acoustic models face a challenging quality-capacity trade-off. We investigate two factors in this regime: encoder context and Mel-spectrogram supervision. A receptive-field-scaling study shows that expanding self-
attention beyond 15 phonemes provides no consistent gains in pitch, energy, or duration prediction. Guided by this finding, we introduce a fixed-receptive-field convolutional encoder that reduces the respective prediction errors by 36.0%, 17.3%, and 3.4%. We further show that directly transferring image-domain gradient-variance supervision restores fine-scale variation but degrades predicted quality, motivating a Mel-specific formulation with axis-specific gradients, overlapping local statistics, and log-domain variance matching. GrainSpeech contains only 264.8K parameters and achieves 17.9x real-time Mel generation on a microcontroller (MCU), while attaining UTMOS scores comparable to substantially larger models with less than 1.5% of their parameters. Source code and demos are available at https://github.com/lab-emi/GrainSpeech.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.DC
Ask the Tool, Don't Guess: Agent Tool Calls Hold Their Progress, and the Serving System Should Read It
Yipeng Liu, Yingqiang Zhang, Feifei Li et al. (+1 more)
An agentic request spends substantial wall-clock time waiting for tools, and its KV cache holds GPU
memory the whole time. Serving systems decide whether that cache stays, leaves, or comes back by guessing how long the tool will run, from the tool's name, its history, a duration declared before the call, or the engine's own occupancy. We show that no estimate fixed before a call starts can know its duration, and such estimates may not even rank the calls. Meanwhile, the running tool already holds the answer, but the
agent stack together with the tool silences it. We propose that tool calls report their progress explicitly while they run, and we measure what that takes. A census of four public
agent corpora finds a readable signal in most tool time once it is revealed, in two strengths: a fraction of the work remaining, or an accurate signal that the end is near. A harness recovers it without changing what the
agent sees, at no measurable cost to the
agent's
benchmark score. At the points where a KV cache decision is made, the reported progress is between several times and an order of magnitude more accurate than the best published predictors, and it stays accurate when the environment changes. Plugged into a production engine through a few small hints, it cuts the p90 time to first
token (TTFT) after a tool call by 20.7% (HBM only) and 20.8% (HBM + DRAM) against LRU, close to an oracle. A serving system should not guess what its tools can tell it.
Read the full abstract ▾
All authors Yipeng Liu, Yingqiang Zhang, Feifei Li, Huanchen Zhang
Artificial Intelligence
Infinite-Parameter LLMs: Generating and Adapting Weights from Live Data
Jinli Hu, Ross M. Clarke, Yichuan Zhang et al. (+1 more)
Scaling laws hold that
language models grow more capable with more parameters and more training data. Mixture-of-Experts (MoE) architectures are a remarkable demonstration of these laws, activating only a fraction of an enormous parameter bank for each
token. But this success is built on static pretraining data — the facts and corrections supplied by users during live interactions are a significant untapped source of potential improvement for a deployed model, but cannot be exploited by conventional architectures whose weights are frozen after training. Instead, this newfound knowledge must be placed in the context (by instruction or
retrieval) and re-read on every request, only to be discarded afterwards. We seek instead to learn from live interactions by dynamically updating model weights. Inspired by MoEs, we propose the Infinite-Parameter LLM. A compact hypernetwork turns the online data into low-rank modulations of a shared base network, so feed-forward weights are generated from live data, not read from static
memory. Whereas existing weight generators are held fixed after reading the context once, we form a
Bayesian belief over the generator's latent state and update it online, such that the effective weights are re-derived as our belief evolves during the session. Although the model's
memory footprint is constant, the feasible space of generated weights is thus effectively infinite. Representing live data in the weights rather than the
prompt amortises compute, frees the context window, persists updates across turns, and can generalise better than in-context use. Our evaluation protocol applies this methodology to in-context learning and
retrieval.
Read the full abstract ▾
All authors Jinli Hu, Ross M. Clarke, Yichuan Zhang, José Miguel Hernández-Lobato
Artificial Intelligence · Cross-listed from cs.CV
Using OCR Heads to Verbalize Image Semantics
Sheridan Feucht, Benno Krojer, Sarah Wang et al. (+3 more)
How do VLMs map from pixels to semantics? To understand this general question, we focus on a narrow one: studying how VLMs perform optical character recognition (OCR). Across four models, we identify
attention heads causally necessary for OCR, and discover that these are in fact general-purpose heads that output interpretable semantic features across all image
tokens. For example, pointing these heads at an image
token containing the word "bike" causes Qwen3-VL-8B to output "bike," but pointing them at a bird wing causes the model to output the
token "feathers." We collapse these heads'
attention weights into a single verbalization lens transformation that reveals interpretable semantic features in hidden states across all layers. When combined with projection to vocabulary space, we can obtain interpretable labels starting from layer 0, showing that image representations are in fact aligned with
language in early layers. We find that we can also use the inverse of this transformation to edit non-word concepts, e.g., replacing a tractor with a revolver in a naturalistic image, providing causal evidence that this subspace is useful for more than just OCR. Our results are an example of how the study of specific mechanisms can shed light on broader interpretability problems.
Read the full abstract ▾
All authors Sheridan Feucht, Benno Krojer, Sarah Wang, Henry Abrahamsen, Byron C. Wallace, David Bau
Artificial Intelligence
Compositional Policy Violations: When Step-Level Compliance Fails In Agentic AI Workflows
Ashwini Kurady, Sri Sai Charith Grandhi, Rajesh Gupta et al. (+1 more)
Agentic workflows now make consequential decisions in regulated settings, and the governance placed around them is almost entirely step-scoped: input-output classifiers, per turn rails, and span-level evaluators. The policies organizations actually hold, such as referral thresholds, authority limits, and review requirements, are properties of the whole execution rather than of any one step. This mismatch admits a failure mode we call a Compositional Policy Violation (CPV): every individual step passes its own check while the composed execution violates the governing policy. A predicate over a single step cannot evaluate a property that step does not determine, so no improvement in the accuracy of the step-scoped monitors detects this class. We define CPVs as the failure of step-level compliance to compose, and present a taxonomy of four types: Authority Creep, Threshold Laundering, Cumulative Sum Violation, and Context Collapse. We show that the correct repair for each class is dictated by where the guarded quantity mutates. We then introduce a provenance-aware runtime architecture that evaluates policies over complete execution traces, recomputing guarded quantities from raw provenance rather than the pipeline's derived representation.
Read the full abstract ▾
All authors Ashwini Kurady, Sri Sai Charith Grandhi, Rajesh Gupta, Sumit Kumar
Artificial Intelligence · Cross-listed from cs.SE
ProgramDistill: From Interactive Web Apps to Verifiable Reference-Guided SWE Tasks
Jeonghye Kim, Minseon Kim, Young Jin Kim et al. (+5 more)
Coding
agents are typically evaluated with desired behavior specified through issues or instructions. In practical web development, however,
agents may need to infer behavior from working software and implement it in an incomplete application. We introduce ProgramDistill, a
benchmark evaluating coding
agents on features discovered through interaction with fully functional reference applications. We build ProgramDistill by factorizing applications into features of different granularities, each associated with replayable behaviors executable via its gold patch. Our pipeline, mine-craft-patch, discovers 1,975 replay-verified behaviors across 26 applications and constructs 4,063 tasks without human intervention. Across nine frontier coding
agents, GPT-6 Astra and Claude Opus 5 achieve 49.2% and 28.8% success on cumulative workflows in full-application reconstruction. In partial-application reconstruction, success falls from 100% to 64.0% and from 96% to 32% as restoration depth increases from 1 to 8. ProgramDistill thus provides a scalable
benchmark with controlled difficulty for evaluating and diagnosing coding
agents, and a natural basis for future curriculum-based training.
Read the full abstract ▾
All authors Jeonghye Kim, Minseon Kim, Young Jin Kim, Matheus Pereira, Marc-Alexandre Côté, Alessandro Sordoni, Xingdi Yuan, Zhengyan Shi
Artificial Intelligence
CERA-MoA: Co-Evolving Routing Mechanisms with Continually Learning LLM Agents
Jiaxuan Jiang, Liyuan He and Zhixuan Fang
Current Mixture-of-
Agents (MoA) paradigms generally treat query routing and
agent fine-tuning as separate processes, limiting their ability to respond to evolving
agent capabilities. This disconnect prevents routing strategies from adapting to evolving
agent capabilities during post-training and prevents
agents from achieving synergistic data-driven specialization. To resolve this, we introduce CERA-MoA (Co-Evolving Router with continually learning
Agents for Mixture-of-
Agents), an iterative reinforcement learning framework where the dynamic router and independent
agent policies co-evolve. We design a predictive familiarity
estimator that leverages mid-layer hidden states to evaluate semantic competence among
agents, avoiding the overhead of full rollouts. Based on these familiarity scores, a cumulative-threshold adaptive routing mechanism dynamically activates a tailored minimal
agent subset, achieving a trade-off between task performance and efficiency. By proactively allocating targeted training samples to
agents based on their evolving competence, CERA-MoA promotes capability differentiation. Extensive experiments across various domains demonstrate that CERA-MoA outperforms state-of-the-art static-
agent routing and fix-workflow fine-tuning baselines.
Read the full abstract ▾
Artificial Intelligence
Version- and Scope-Aware Question Answering over Normative Documents: A Deployed System and an End-to-End Evaluation at Production Scale
Liuyin Wang, Shuaipeng Jin, Jiwei Shi et al. (+1 more)
Correctly answering a question grounded in normative documents often depends on information outside any single passage: whether the retrieved document is the version currently in force; whether it applies to the jurisdiction, subject (such as an institution or applicant), and date at issue; and whether each normative claim can be traced to its supporting source text. Hosted
retrieval services have substantially lowered the engineering cost of building an initial system over such corpora, making "upload the documents and ask" a common default. We evaluate this default on approximately 73,000 candidate normative documents supplied to a production deployment. The evaluation uses a stratified sample of 200 questions from our published
benchmark, with a gold source document for every question; the released sampling rule reads no system outputs or scores. We compare the hosted service with a governed system that resolves version and scope through explicit rules before generation. The governed system scored 97.7 overall, while the hosted service scored 88.1, a gap of 9.6 points computed from unrounded means. The question set, the answer text evaluated for both systems, the scores, and the scripts used to reproduce the reported
benchmark statistics are public. The governed configuration has operated as a commercial product since January 2026 and serves 1,126 registered users; named customer organizations include Zhipu AI and Lecheng Health. By mid-April 2026, it had reached roughly 100,000 calls per workday.
Read the full abstract ▾
All authors Liuyin Wang, Shuaipeng Jin, Jiwei Shi, Jensen Hsu
Artificial Intelligence · Cross-listed from cs.CL
A Scalable Framework for Automated NER Annotation Correction in Low-Resource Languages
Toqeer Ehsan and Thamar Solorio
Poor quality or noisy annotations in Named Entity Recognition (NER), as in any other NLP task, make it challenging to achieve state-of-the-art performance. In this paper, we present a multi-step framework to enhance the annotation quality of NER
datasets by employing automated techniques. We propose a frequency-based iterative approach that leverages self-training and a dual-threshold mechanism to enhance inference confidence. Experimental evaluations on different NER
datasets demonstrate significant improvements in NER performance with respect to the original
datasets. This work further explores the potential of generative Large
Language Models (LLMs) to perform NER for low-resource
languages.
Read the full abstract ▾
Artificial Intelligence
Clueing up LLMs with Tool-Augmented Deductive Reasoning
Rebecca Ansell and Autumn Toney-Wails
Despite recent advances in large
language models (LLMs), performing logically consistent deductive
reasoning over extended interactions remains challenging. Tasks that require integrating evidence across multiple
reasoning steps, maintaining consistency with prior inferences, and updating beliefs under new constraints can surface limitations in current models while providing a useful testbed for evaluating
reasoning enhancements. In this paper, we implement a text-based, multi-
agent version of the classic board game Clue as an environment to evaluate multi-step, agentic deductive
reasoning. In this setting,
agents must infer hidden information from a sequence of observations, maintain consistency across turns, and reason over an evolving set of logical constraints. We instantiate six LLM-based
agents (GPT-4o-mini and Gemini-2.5-Flash) as players that engage in turn-based gameplay; using three
agents per model family, we establish baseline performance across repeated games. We then introduce a tool-augmented approach in which a structured possibility
matrix converts implicit game state from generated
reasoning logs into an explicit representation of remaining possibilities. The possibility
matrix encodes extended-turn
memory and deductive constraints, offloading these tasks from the
agent. We compare this approach against the baseline to evaluate how tool augmentation supports
reasoning quality and task success for autonomous
agents in a strategic
reasoning environment.
Read the full abstract ▾
Artificial Intelligence
Which LLM is Best for Translating Natural Language Goals to PDDL
Tomas Balyo, Lukas Chrpa and G. Michael Youngblood
Bridging the gap between human intent and machine execution remains a challenge in automated planning, where expressing goals in formal
languages like PDDL restricts accessibility to non-experts. This paper empirically evaluates whether current Large
Language Models (LLMs) can reliably translate natural
language testing goals, written in informal
language by
video game testers, into well-formed PDDL targets suitable for classical planning. We present a carefully designed
prompt template, integrating insights from iterative experimentation, aimed at maximizing both accuracy and response coherence from multiple state-of-the-art LLMs. Six contemporary models are systematically assessed on correctness, speed, and error tendencies using real-world, domain-specific
benchmarks. All models demonstrate high correctness, exceeding 92%, with Gemini 2.5 Flash achieving the highest accuracy at 96% and the lowest incidence of false positives, while GPT-4.1 leads in response speed. Despite these advances, critical distinctions exist in model performance, and occasional failures arise from
language ambiguity and limitations in domain representation. Our analysis underscores both the significant progress and ongoing gaps in enabling LLMs to act as robust bridges between natural
language objectives and automated planning pipelines.
Read the full abstract ▾
Artificial Intelligence
Beyond Truncation: Rethinking LLM Decoding as Ensemble Pruning
Dunyao Xue, Chengshuo Du, Zhengbo Wang et al. (+2 more)
We introduce Mahalanobis-Ensemble Decoding (ME-Decoding), a novel Large
Language Model (LLM) decoding framework that frames candidate
token selection as ensemble
pruning. Existing selection strategies rely predominantly on scalar probabilities, ignoring
geometric semantic relationships and causing candidate redundancy. Meanwhile, current geometry-aware methods often require complex optimization or directly reweighting the original
token probabilities, leading to significant computational overhead or inference instability. To address this, we formulate decoding as a subset optimization problem using a Mahalanobis distance-driven objective to enhance semantic diversity while preserving high probabilities. Specifically, we dynamically discount redundant generation paths using a
token similarity
matrix, constructed via an adaptive-bandwidth kernel over
token embeddings. We further devise an efficient greedy selection
algorithm with near-linear complexity in the candidate size under early stopping, while establishing its theoretical approximation guarantees. This renders ME-Decoding a robust, plug-and-play module with negligible inference overhead. Extensive experiments across diverse
reasoning and generation tasks demonstrate that our method consistently achieves strong performance.
Read the full abstract ▾
All authors Dunyao Xue, Chengshuo Du, Zhengbo Wang, Wenlin Dai, Cheng Meng
Artificial Intelligence · Cross-listed from cs.LG
Rethinking Critic Learning in PPO: Understanding and Mitigating Value Flattening
Yizhuo Li, Jianhao Yan, Yun Luo et al. (+9 more)
In reinforcement learning for large
language models, Proximal Policy Optimization (PPO) commonly uses a critic to estimate state values and reduce the variance of policy updates. However, we uncover a systematic failure mode in PPO critics, which we call Value Flattening: state values, estimated from multiple Monte Carlo continuations, change sharply across intermediate states while critic predictions remain comparatively flat. We further observe this phenomenon in a controlled FrozenLake environment and find that it becomes more pronounced as the state space grows. Our theoretical and empirical analyses relate Value Flattening to an implicit variance penalty in the critic loss and redundant updates from temporally correlated states with similar gradients. Motivated by these findings, we introduce SParse Proximal Policy Optimization (SP$^3$O), which applies the value loss to only a few well-separated states in each response to mitigate both effects. Experiments on Qwen3-Base show that SP$^3$O with only three states supervised per response can mitigate Value Flattening and consistently improve the learned policy across model sizes and evaluation suites. Together, our results identify Value Flattening as an important yet overlooked failure mode of critic learning in standard PPO and show that a simple sparse supervision strategy can mitigate it.
Read the full abstract ▾
All authors Yizhuo Li, Jianhao Yan, Yun Luo, Zhi Wang, Futing Wang, Rong-Xi Tan, Kanghui Tian, Ganqu Cui, Ning Ding, Peilin Zhao, Yafu Li, Yu Cheng
Artificial Intelligence · Cross-listed from cs.CR
Echo: Learning-based Matching Decompilation using Trusted Back Translation
Jun Bi, Xiangxin Fang, Aarsh Chaube et al. (+3 more)
Neural decompilers can recover readable and recompilable source code from binaries, but their predictions remain difficult to trust. Matching decompilation addresses this problem by searching for source code whose recompiled assembly exactly matches the target, providing stronger evidence of correctness. However, exact matching remains challenging for optimized binaries under unknown compilation configurations. We present Echo, a matching decompilation system based on trusted back-translation. Our key insight is to use compilation not only for verification, but also as trusted feedback to guide iterative search. Echo first uses a domain-specific model to generate candidate programs and compilation configurations. It recompiles these candidates, measures assembly-level similarity, and synthesizes promising code-configuration pairs. Remaining mismatches are then progressively repaired using rule-based rewriting, neural refinement, and
reasoning-based refinement. We evaluate Echo on function-level
benchmarks and the Mirai malware binary. Compared with the strongest baseline, Echo produces 2.43x more exact matches on average and achieves the highest structural similarity to ground-truth source code. On Mirai, Echo matches 2.75x and 7.4x as many functions as GPT-5.6 and Codex, respectively.
Read the full abstract ▾
All authors Jun Bi, Xiangxin Fang, Aarsh Chaube, José Wesley De Souza Magalhães, Rodrigo C. O. Rocha, Michael O'Boyle
Artificial Intelligence · Cross-listed from cs.CV
Generalist-Specialist Mixture-of-Experts for Rare Pathology Detection in Multimodal Imaging
Johannes Kaiser, Florian Braunmiller, Daniel Rückert et al. (+1 more)
AI models for
multimodal medical imaging must balance modality-specific specialization with cross-modal shared representations, a trade-off that pure Mixture-of-Experts (MoE) architectures currently fail to satisfy. Expert-based routing improves in-domain learning but may sacrifice cross-modal signals, which appear particularly important for rare (low-prevalence) pathologies in our experiments. To resolve this, we introduce Generalist-Specialist-MoE (GS-MoE), a two-branch (MoE) architecture that couples a cross-modal generalist model with distinct modality-specific specialists (experts) via domain-constrained feature fusion. On RadImageNet (1.35M images, 165 pathologies, three modalities), GS-MoE recovers detection of six low-prevalence pathologies on which every baseline scores F1 $=$ 0, with per-class gains up to +0.60 F1. It attains this while even slightly exceeding dense and specialist-only MoE aggregate baselines (MCC 0.770), while using ${\sim}53\%$ fewer active parameters at inference than the strongest investigated dense model.
Read the full abstract ▾
All authors Johannes Kaiser, Florian Braunmiller, Daniel Rückert, Georgios Kaissis
Artificial Intelligence
The Uneven Impact of Generative AI on Student Learning: Examining the Roles of Reliance, Evaluation Literacy, and Course Policy in AI-related Courses
Lydia Manikonda, Mei Si, Sirajam Munira et al. (+2 more)
Generative artificial intelligence (GenAI) is changing how students learn, yet the roles of course context, cognitive reliance, evaluation literacy, and early reliance remain underexplored. Using survey responses from 118 students across 12 AI-related courses at our institution, we examined differences in GenAI use and perceived learning experiences. We identified four user clusters: high-use students reporting many benefits, light users reporting less reliance and fewer benefits, and two moderate-use groups reporting different levels of benefit. We also found significant differences between free- and premium-version users, single- and multiple-tool users, and students experiencing different instructor policies. In multivariable regression models, academic benefit was associated with early reliance and academic task support; positive impact was associated with cognitive reliance, academic task support, confidence in GenAI reliability, and instructor policy; and negative impact was associated with early reliance and attitudinal change. The association between early reliance and negative impact became stronger as evaluation literacy increased. Finally, perceptions of GenAI-enhanced learning appear to reflect cognitive, performance, and self-efficacy benefits, while concerns about stress and diminished critical thinking are associated with lower perceived learning benefits. These findings suggest that institutions need better policies to address such inequities so that institutions can enable students to benefit from increasingly capable AI systems.
Read the full abstract ▾
All authors Lydia Manikonda, Mei Si, Sirajam Munira, Oshani Seneviratne, Kristin Bennett
Artificial Intelligence · Cross-listed from cs.SD
Beyond EER: Multi-Dimensional Evaluation of Information Leakage in Speaker De-Identification
Seungmin Seo, Oleg Aulov, P. Jonathon Phillips et al. (+2 more)
Speaker de-identification (SDID) aims to preserve
privacy by concealing speaker identity while maintaining
speech utility. However, current evaluations often reduce
privacy to a single dimension - biometric verification performance - typically measured by Equal Error Rate (EER). This narrow focus ignores critical leakage channels, such as soft biometric inference,
embedding-level re-identification, and structural template similarity, which threaten the unlinkability and irreversibility of biometric references. We propose a holistic evaluation framework across five complementary metrics: (i) EER, (ii) soft biometric leakage score , (iii) cumulative match characteristic re-identification analysis, (iv) canonical correlation analysis and Procrustes
embedding alignment, and (v) intelligibility via word error rate and semantic similarity. Evaluating five SDID systems from the IARPA ARTS program, we demonstrate that these metrics capture independent dimensions of information leakage. Our results indicate that reliance on a single metric can misrepresent the
privacy properties of an SDID system.
Read the full abstract ▾
All authors Seungmin Seo, Oleg Aulov, P. Jonathon Phillips, Kevin Mangold, Jonathan Eskin