arXiv:2505.11470v3 Announce Type: replace
Abstract: We introduce two reference-free metrics for quality evaluation of taxonomies in the absence of labels. The first metric evaluates robustness by calculating the correlation between semantic and taxonomic similarity, addressing error types not considered by existing metrics. The second uses Natural Language Inference to assess logical adequacy. Both metrics are tested on five taxonomies and are shown to correlate well with F1 against ground truth taxonomies. We further demonstrate that our metrics can predict downstream performance in hierarchical classification when used with label hierarchies.
Science Journals
arXiv:2505.23131v2 Announce Type: replace
Abstract: We study the problem of assigning operations in a dataflow graph to devices to minimize execution time in a work-conserving system, with emphasis on complex machine learning workloads. Prior learning-based methods often struggle due to three key limitations: (1) reliance on bulk-synchronous systems like TensorFlow, which under-utilize devices due to barrier synchronization; (2) lack of awareness of the scheduling mechanism of underlying systems when designing learning-based methods; and (3) exclusive dependence on reinforcement learning, ignoring the structure of effective heuristics designed by experts. In this paper, we propose Doppler, a three-stage framework for training dual-policy networks consisting of 1) a $\mathsf{SEL}$ policy for selecting operations and 2) a $\mathsf{PLC}$ policy for placing chosen operations on devices. Our experiments show that Doppler outperforms all baseline methods across tasks by reducing system execution time and additionally demonstrates sampling efficiency by reducing per-episode training time.
arXiv:2507.01548v3 Announce Type: replace
Abstract: This paper explores how older migrants in urban China can record stories that everyday language and design often miss. We ran two co-creation workshops with 10 elders. Activities combined oral storytelling, facilitator-mediated AI assistance, and hand-making. Large language models proposed candidate glyphs through a facilitator. Participants crafted new Hanzi to hold their stories. The resulting characters served as memory anchors for later sharing and retelling. Our interpretive analysis shows heterogeneity and adaptive capacity among participants. Participants experienced AI as a creative initiator that lowered barriers to expression and making, especially for those with lower digital literacy. The work challenges homogenizing assumptions about older adults and the presumption of uniform capacities and needs. We contribute a workshop framework that positions AI as a backstage facilitator. We also offer insights on engaging older migrants as sources of community memory and situated cultural knowledge within inclusive urban systems.
arXiv:2511.02748v2 Announce Type: replace
Abstract: We argue that sixth-generation (6G) intelligence is not fluent token prediction but the capacity to imagine and choose -- to simulate future scenarios, weigh trade-offs, and act with calibrated uncertainty. We reframe open radio access network (O-RAN) near-real-time (Near-RT) control via counterfactual dynamics and a world modeling (WM) paradigm that learns an action-conditioned generative state space. This enables quantitative "what-if" forecasting beyond large language models (LLMs) as the primary modeling primitive. Actions such as physical resource blocks (PRBs) are treated as first-class control inputs in a causal world model, and both aleatoric and epistemic uncertainty are modeled for prediction and what-if analysis. An agentic, model predictive control (MPC)-based cross-entropy method (CEM) planner operates over short horizons, using prior-mean rollouts within data-driven PRB bounds to maximize a deterministic reward. The model couples multi-scale structured state-space mixtures (MS3M) with a compact stochastic latent to form WM-MS3M, summarizing key performance indicators (KPIs) histories and predicting next-step KPIs under hypothetical PRB sequences. On realistic O-RAN traces, WM-MS3M cuts mean absolute error (MAE) by 1.69% versus MS3M with 32% fewer parameters and similar latency, and achieves 35-80% lower root mean squared error (RMSE) than attention/hybrid baselines with 2.3-4.1x faster inference, enabling rare-event simulation and offline policy screening.
arXiv:2511.05949v2 Announce Type: replace
Abstract: While stereo matching has achieved maturity for 0D point and 1D line primitives, establishing correspondences for 2D polygons remains largely unexplored due to challenges including disparity discontinuity, scale variation, training dependency, and poor generalization, limiting downstream tasks such as pose estimation and 3D reconstruction. To address these issues, we are the first to propose a Zero-shot Polygon Matching paradigm with Pre-trained Models (i.e., Z(PM)2), which combines learned features and handcrafted geometric constraints through plug-and-play modules, extending matching from 0D/1D primitives to 2D polygons. The pipeline comprises three core stages: Firstly, detector leverages the pre-trained segment anything model to vectorize segmentation masks into graph-structured polygons integrating geometry and texture; Secondly, global matcher uses bidirectional-pyramid and multi-geometric constraints to handle viewpoint variation; Thirdly, local matcher leverages local-holistic bipartite graph optimization to resolve disparity discontinuity and topological inconsistency. Moreover, we develop polygon-matching-guided pose estimation using correspondences to obtain well-distributed, low-redundancy homologous points, and pioneer the polygon cloud concept with an optimal surface generation method, producing structurally complete and semantically rich 3D representations beyond point and line clouds. Since no polygon matching methods from stereo imagery are available for direct comparison, we selected state-of-the-art (SoTA) methods close to this task as baselines. Extensive experiments on five challenging datasets (ISPRS, KITTI, ScanNet, SceneFlow, DTU) show Z(PM)2 achieves a 68.60% matching area score, outperforming MESA by approximately 32% and ranking first in area-level pose estimation, with competitive speed and strong zero-shot generalization without any training requirement.
arXiv:2511.10763v2 Announce Type: replace
Abstract: Uncrewed Aerial Vehicles (UAVs) serving as Aerial Base Stations (ABSs) are expected to extend 6G millimeter-Wave (mmWave) coverage and improve link reliability in urban areas. However, UAV-based Air-to-Ground (A2G) channels are highly dependent on height and urban geometry. This paper proposes an ABS height-dependent mmWave channel model and investigates whether urban geometry, beyond the standard built-up parameters, significantly affects LoS probability (PLoS) and Large-Scale Fading (LSF). Using MATLAB ray tracing at 26 GHz, we simulate approximately 10K city realizations for four urban layouts that share identical built-up parameters but differ in their spatial organization. We extract elevation-based PLoS using a sigmoid model and derive height-dependent Path-Loss Exponents (PLEs) and shadow-fading trends using exponential fits. Results show that PLE for Non-Line-of-Sight (NLoS) decreases toward 2.5-3 at high altitudes, Line-of-Sight (LoS) PLE remains near 2, and shadow fading reduces with height. We also find that geometric layout introduces a modest but consistent change in PLE (+/- 0.2), even when built-up parameters are fixed. The proposed unified model aligns well with ray-tracing statistics and offers a practical, height-dependent LSF model suitable for ABS planning in complex urban scenarios.
arXiv:2512.23292v5 Announce Type: replace
Abstract: The prevailing paradigm in AI for physical systems: scaling general-purpose foundation models toward universal multimodal reasoning, confronts a barrier at the control interface. Frontier vision-language models achieve only 50-53% accuracy on basic quantitative physics tasks, behaving as approximate guessers that preserve semantic plausibility while violating physical constraints. Safety-critical control demands outcome-space guarantees over executed actions, not parameter-space imitation. Here we present a pathway toward domain-specific foundation models through compact language models operating as Agentic Physical AI: policy optimization driven by physics-based simulator validation rather than perceptual inference. We train a 360M-parameter model on synthetic nuclear reactor scenarios scaled from 10^3 to 10^5 examples. Scaling produces strong, regime-dependent reliability gains under nominal simulated conditions, with variance collapse of approximately 500x and elimination of >10% terminal-power excursions on the sampled distribution. Despite balanced exposure to four actuation families, the model concentrates 95% of runtime execution on a single-bank strategy, without reinforcement learning or reward engineering. Representations transfer across simulators without architectural change. We position the system as a candidate decision component within a verification, monitoring, and defense-in-depth architecture, not as a stand-alone safety solution: the demonstrated behavior speaks to closed-loop reliability on a single-step task in simulation and does not yet address off-nominal operation, sensor faults, or uncertainty quantification.
arXiv:2512.23294v2 Announce Type: replace
Abstract: Semantic communications (SemCom), as one of the key technologies for 6G, is shifting networks from bit transmission to semantic information exchange. On this basis, introducing agentic artificial intelligence (AI) with perception, memory, reasoning, and action capabilities provides a practicable path to intelligent communications. This paper provides a systematic exposition of how agentic AI empowers SemCom from the perspectives of research foundations, system architecture, and application scenarios. We first provide a comprehensive review of existing studies by agent types, covering embedded agents, large language model (LLM)/large vision model (LVM) agents, and reinforcement learning (RL) agents. Additionally, we propose a unified agentic AI-enhanced SemCom framework covering the application layer, the semantic layer, and the cloud-edge collaboration layer, forming a closed loop from intent to encoding to transmission to decoding to action to evaluation. We also present several typical scenarios, including multi-vehicle collaborative perception, multi-robot cooperative rescue, and agentic operations for intellicise (intelligent and concise) networks. Furthermore, we introduce an agentic knowledge base (KB)-based joint source-channel coding case study, AKB-JSCC, where the source KB and channel KB are built by LLM/LVM agents and RL agents, respectively. Experimental results show that AKB-JSCC achieves higher information reconstruction quality under different channel conditions. Finally, we discuss future evolution and research directions, providing a reference for portable, verifiable, and controllable research and deployment of agentic SemCom.
arXiv:2601.04791v4 Announce Type: replace
Abstract: While latent diffusion models (LDMs) have emerged as powerful priors for inverse problems, existing LDM-based solvers frequently suffer from instability. In this work, we first identify the instability as a discrepancy between the solver dynamics and stable reverse diffusion dynamics learned by the diffusion model, and show that reducing this gap stabilizes the solver. Building on this, we introduce \textit{Measurement-Consistent Langevin Corrector (MCLC)}, a theoretically grounded plug-and-play stabilization module that remedies the LDM-based inverse problem solvers through measurement-consistent Langevin updates. Compared to prior approaches that rely on linear manifold assumptions, which often fail to hold in latent space, MCLC provides a principled stabilization mechanism, leading to more stable and reliable behavior in latent space.
arXiv:2601.19542v2 Announce Type: replace
Abstract: This paper presents an axisymmetric Galerkin boundary element method (BEM) for modeling eddy-current interactions between excitation coils and conductive objects. The formulation derives boundary integral equations from the Stratton-Chu representation for the azimuthal component of the vector potential in both air and conductive regions. The central contribution is a unified regularization framework for the two-dimensional (2D) singular integrals arising in Galerkin BEM. This framework handles both logarithmic and Cauchy singularities through a common set of integral transformations, eliminating the need for case-by-case analytical singularity extraction and enabling straightforward numerical quadrature. The regularization and quadrature stability are proved and verified numerically. The method is validated on several representative axisymmetric geometries, including cylindrical, conical, and spherical shells. Numerical experiments demonstrate consistently high accuracy and computational efficiency over the tested frequency interval and coil lift-off distances. The results confirm that the proposed axisymmetric Galerkin BEM, combined with the integral transformation technique, provides a robust and efficient framework for axisymmetric eddy-current nondestructive evaluation.
arXiv:2602.00541v2 Announce Type: replace
Abstract: Clinical events captured in Electronic Health Records (EHR) are irregularly sampled and may consist of a mixture of discrete events and numerical measurements, such as laboratory values or treatment dosages. The sequential nature of EHR, analogous to natural language, has motivated the use of next-token prediction to train prior EHR Foundation Models (FMs) over events. However, this training fails to capture the full structure of EHR. When a given event occurs must be captured, but the event value (abnormal lab) also modulates the likelihood of other clinical events. Most existing EHR FMs do not jointly model this likelihood and are unable to capture the full observation process, impacting downstream capabilities. We propose ORA, a marked time-to-event pretraining objective that jointly models event timing and associated measurements. Across multiple datasets, downstream tasks, and model backbones, this objective consistently yields more generalizable representations than next-token prediction and pretraining losses that ignore continuous measurements. Importantly, the proposed objective yields improvements beyond traditional classification evaluation, including better regression and time-to-event prediction. Beyond introducing a new family of FMs, our ablations suggest a broader takeaway: pretraining objectives that account for EHR structure are critical for expanding downstream capabilities and generalizability.
arXiv:2602.01740v3 Announce Type: replace
Abstract: Video language models (Video-LLMs) are prone to hallucinations, generating plausible but ungrounded content when visual evidence is weak, ambiguous, or biased. Existing methods, such as contrastive decoding (CD), rely on random perturbations to construct contrastive data for hallucination mitigation, but often fail to target the visual cues that drive hallucination or align with model weaknesses. We propose Model-Aware Counterfactual Data based Contrastive Decoding (MACD), an inference strategy that combines model-guided counterfactual construction with contrastive decoding. MACD uses the Video-LLM's own feedback to identify object regions most responsible for hallucination, generating targeted object-level counterfactual inputs rather than arbitrary frame or temporal modifications. These counterfactual inputs are integrated into CD to enforce evidence-grounded token selection during decoding. Experiments on EventHallusion, MVBench, Perception-test, and Video-MME show that MACD consistently reduces hallucination while maintaining or improving task accuracy across diverse Video-LLMs, including Qwen and InternVL, with especially strong gains in scenarios involving small, occluded, or co-occurring objects.
arXiv:2602.08857v2 Announce Type: replace
Abstract: Recent work has shown that the computations of Transformers can be simulated in the RASP family of programming languages. These findings have enabled improved understanding of the expressive capacity and generalization abilities of Transformers. In particular, Transformers have been suggested to length-generalize exactly on problems that have simple RASP programs. However, it remains open whether trained models actually implement simple interpretable programs. In this paper, we present a general method to extract such programs from trained Transformers. The idea is to faithfully re-parameterize a Transformer as a RASP program and then apply causal interventions to discover a small sufficient sub-program. In experiments on small Transformers trained on algorithmic and formal language tasks, we show that our method often recovers simple and interpretable RASP programs from length-generalizing transformers. Our results provide the most direct evidence so far that Transformers internally implement simple RASP programs.
arXiv:2604.08168v2 Announce Type: replace
Abstract: Vision-language-action (VLA) models have advanced robot manipulation through large-scale pretraining, but real-world deployment remains challenging due to partial observability and delayed feedback. Reinforcement learning addresses this via value functions, which assess task progress and guide policy improvement. However, existing value models built on vision-language models (VLMs) struggle to capture temporal dynamics and physical interactions, undermining reliable value estimation in long-horizon tasks. In this paper, we propose ViVa, a video-generative value model that repurposes a pretrained video generator to jointly predict future proprioception and a scalar value. By grounding value estimation in anticipated embodiment dynamics, ViVa leverages spatiotemporal priors to intrinsically couple value with foresight beyond static snapshots. ViVa achieves state-of-the-art results in metric-based evaluation across three tasks, producing reliable value signals that accurately track task progress and detect execution errors. Integrated into RECAP, it achieves an average success rate of 80%, highlighting the promise of video-generative models for value estimation.
arXiv:2604.15290v4 Announce Type: replace
Abstract: A promising approach to unifying functional and imperative programming paradigms is to localize mutation using linear or affine types. Haskell, a purely functional language, was recently extended with linear types by Bernardy et al., in the name of Linear Haskell. However, it remained unknown whether such a pure language could safely support non-local borrowing in the style of Rust, where each borrower can be freely split and dropped without direct communication of ownership back to the lender.
We answer this question affirmatively with Pure Borrow, a novel framework that realizes Rust-style borrowing in Linear Haskell with purity. Notably, it features parallel state mutation with affine mutable references inside pure computation, unlike the IO and ST monads and existing Linear Haskell APIs. It also enjoys purity, lazy evaluation, first-class polymorphism and leak freedom, unlike Rust. We implement Pure Borrow simply as a library in Linear Haskell and demonstrate its power with a case study in parallel computing. We formalize the core of Pure Borrow and build a metatheory that works toward establishing safety, leak freedom and confluence, with a new, history-based model of borrowing.
arXiv:2605.30432v2 Announce Type: replace-cross
Abstract: Social systems consist of networks of individuals who influence one another through social interactions. Studying how processes evolve on these networks can help us better understand patterns of social behavior. We study a system that couples online and offline social activity and investigate how to learn effective models directly from data using Weak Form Sparse Identification of Nonlinear Dynamics (WSINDy), a method for discovering governing equations. We assess learning performance using data generated by a mean-field approximation model of a stochastic interaction process on networks and test how accurately the system can be recovered under different noise levels. Our results show that using more trajectories improves accuracy when noise is high, but only a small number of additional trajectories is needed to gain most of the benefit, with little improvement beyond that. We also learn effective ODE models from averaged stochastic data on networks. When traditional mean-field approximations fail, identifying continuum ODEs directly from stochastic processes yields efficient models that better match the data and provide deeper insight into the underlying dynamics.
arXiv:2606.06502v1 Announce Type: new
Abstract: As large language models (LLMs) are increasingly trained on scraped web corpora without authorisation, content owners require forensic methods to prove that their documents were included in a model's training set. We propose \textbf{SIGIL} (\textbf{S}ubtle \textbf{I}njection for \textbf{G}round-truth \textbf{I}nference of \textbf{L}LM training data), a framework that embeds imperceptible \emph{canary sequences} into protected text and code such that any LLM trained on those documents exhibits statistically detectable behavioural signatures when probed with targeted queries.
SIGIL defines five canary strategies -- lexical-rare, lexical-phrase, syntactic, semantic, and code-pattern -- and a \emph{Membership Inference Score} (MIS) grounded in the Neyman-Pearson hypothesis testing framework with formal false-positive rate (FPR) control. Simulator parameters are calibrated against the empirical membership inference literature, yielding realistic heterogeneous results across $36{,}000$ trials: overall AUC $= 0.892$, rising from $0.831$ at $0.1\%$ injection to $0.947$ at $10\%$. Detection rates range from $33\%$ to $78\%$ across model-size and injection-rate conditions. Code Pattern canaries achieve the highest AUC ($0.903$, Cohen's $d = 1.84$); Syntactic Structure the lowest ($0.875$, $d = 1.63$). All four experimental factors -- injection rate, model size, canary strategy, and mixture ratio -- have significant independent effects on MIS ($p < 0.001$). SIGIL maintains AUC $> 0.86$ even under $100\%$ paraphrasing ($\text{AUC} = 0.864$), confirming robustness through semantic leakage that survives surface-level rewriting.
arXiv:2606.06504v1 Announce Type: new
Abstract: The enthesis is the insertion site of tendon into bone and exhibits a high mechanical durability despite the large mismatch in material properties between the two tissues. This durability stems from gradients in composition, structure and organization on multiple hierarchical length scales. Despite extensive research on enthesis structure and mechanics, the local deformation mechanisms are poorly understood. Synchrotron scanning small- and wide-angle X-ray scattering was combined with in situ tensile testing of the mouse Achilles tendon enthesis to extensively map the mechanical response of the collagen fibrils and molecules as well as the hydroxyapatite mineral particles and crystals. Gradients in nano- and molecular scale structure and a stronger and more immediate deformation response towards the interface compared to further away were observed in both the soft and mineralized tissue. The strain decreased progressively across hierarchical levels; with an applied tissue strain of 20% the nanoscale fibrils were strained by ~1-2%, the collagen molecules by ~0.5% and the hydroxyapatite crystals by ~0.05%, thus following an approximate ratio of 1 : 0.1 : 0.01 : 0.001. These results show that load transfer across the enthesis is both spatially heterogeneous and hierarchy-dependent. This indicates that the graded attachment accommodates deformation through region-specific load sharing and hierarchical strain partitioning, consistent with a contribution from dissipation within the non-collagenous matrix. In doing so, the enthesis can mitigate stress concentrations and maintain mechanical integrity across the tendon-to-bone transition.
arXiv:2606.06647v1 Announce Type: new
Abstract: Objective. EEG foundation models (FMs) report strong accuracy on clinical resting-state EEG. However, high accuracy under subject-disjoint cross-validation remains ambiguous: it can reflect a genuine clinical biomarker, or subject-identity features that correlate with the label. We name this the Identity Trap and ask whether it can be diagnosed at the representation level before fine-tuning.
Approach. We propose FMScope, a frozen-representation protocol packaging five diagnostics: variance decomposition, subject-axis erasure, aperiodic 1/f ablation, layer-wise label probing, and within-subject direction consistency. We apply it to three pretrained FMs (LaBraM, CBraMod, REVE) across four datasets in a 2x2 layout: subject relation of label x presence of a consensus cross-subject EEG marker.
Main results. (i) The Identity Trap is universal: frozen subject-variance is 13-89x a random null in 12/12 pairs, rising in all 12 under fine-tuning (+10 to +63 pp). This dominance is a removable linear axis: erasing it improves label decoding where the label varies within subject (+6 to +12 pp in primary cells; +4 to +27 pp across external cohorts). (ii) Aperiodic 1/f is one subject carrier: removing it drops the subject probe by 9-19 pp on LaBraM and CBraMod. REVE saturates subject identity without measurable aperiodic dependence. (iii) Fine-tuning amplifies label-variance only in cells with a literature-established cross-subject marker.
Significance. The Identity Trap is a physically-grounded instance of shortcut learning: the preferred cue has a measurable physiological component, and subject-disjoint splitting alone cannot rule it out. FMScope separates gains reflecting a biological marker from those reflecting subject identity.
arXiv:2606.06662v1 Announce Type: new
Abstract: Modern software development relies on CI/CD pipelines to automate testing, building, and deployment operations. Configuring DevOps pipelines is challenging and time-consuming, as developers must understand platform-specific syntax and manually create configuration files. This complexity can lead to configuration errors and reduced productivity, especially for developers with limited DevOps experience. This paper introduces the AutoPipelineAI system, which generates CI/CD pipeline configurations using natural language descriptions. The proposed solution uses large language models (LLMs) to translate developer intent, analyze repository structures, and create specific pipeline scripts for environments like GitHub Actions and GitLab CI/CD. It integrates repository-aware analysis, automated validation systems, and a feedback mechanism that confirms the accuracy and usability of the created pipelines. We present the system architecture, its implementation, and an assessment framework designed to measure generation precision, configuration validity, and reduction in setup effort compared to manual pipeline creation. AutoPipelineAI illustrates how LLMs can simplify the complexity of DevOps configuration and enhance developer access to continuous delivery methods. Evaluation results provide early evidence that repository-aware, natural-language-driven CI/CD generation is a viable and promising paradigm for reducing the complexity of DevOps configuration and enabling more accessible software delivery automation.
arXiv:2606.06666v1 Announce Type: new
Abstract: Deepfake detection systems achieve near-perfect accuracy on benchmarks, yet forensic deployment demands reliable prediction uncertainty. Existing uncertainty quantification (UQ) methods rely on single sources and ignore that optimal uncertainty composition varies across architectures. We propose Correlation-Optimized Fusion (COF), an architecture-adaptive framework that fuses five complementary uncertainty sources -- epistemic, aleatoric, calibration, conformal, and distributional -- by maximizing Pearson correlation between fused uncertainty scores and prediction errors via constrained optimization on the probability simplex. COF requires no model modifications and only 42 s of weight optimization, compared to 20--45 h for a 5-model Deep Ensemble. Evaluation across eleven architectures on FaceForensics++ reveals a fundamental trade-off: under matched train/evaluation protocol, non-linear methods achieve approximately 5--6% higher in-domain correlation than COF (mean r = 0.438), but this reverses under distribution shift. On CelebDF, COF outperforms Random Forest in 9/11 architectures with up to 7.3x higher correlation (MaxViT-B: r = 0.249 vs. 0.034); RF degrades 85% cross-domain to r = 0.071, whereas COF retains substantially more signal (74% drop to r = 0.116). Cross-dataset evaluation on CelebDF and DFDC reveals catastrophic generalization failure across all methods: in-domain correlations of 0.41--0.47 collapse to near-zero externally (mean degradation 90.7%), with seven of eleven architectures exhibiting uncertainty inversion. These results establish COF as a practical, interpretable framework for controlled-distribution deployment and identify domain-adaptive UQ as the central open challenge for forensic deployment.
arXiv:2606.06722v1 Announce Type: new
Abstract: The training of neural networks often entails objective functions that are not globally $L$-smooth. For these functions, it is both theoretically and practically difficult to reply to the question: what is the largest possible step size that ensures the convergence of gradient descent (GD)? We address this longstanding open question in deep learning by providing a unifying definition of "large" step sizes that requires only local Lipschitz (or even H\"older) continuity of the gradient. We design first-order adaptive methods that provably yield large step sizes and show that they operate at the edge of stability (EoS) right from the start of the training. In particular, the loss decreases nonmonotonically and the product between the step size and sharpness, i.e., the largest eigenvalue of the Hessian, stays above the EoS threshold of 2 throughout training. Using our method, we are also able to minimize the sharpness all the way down to its global minimum. Contrary to expectation, we find that encountering globally-flat regions too early in the training may both slow down convergence and jeopardize the generalization ability of the network. Exploiting a self-stabilization argument, we allow GD to enter slightly sharper valleys and turn unsuccessful training runs into very successful ones.
arXiv:2606.06724v1 Announce Type: new
Abstract: Representative data is fundamental in machine learning, as limited data hinders generalisation. Collecting sufficient real-world samples is often infeasible. Synthetic data generation offers a practical solution, but only if the generated data faithfully reflects the structure of real observations. In this paper, a method for generating synthetic regression datasets that structurally resemble physics equations from a given equation corpus is presented. The approach uses a Bayesian Probabilistic Context-Free Grammar to capture the underlying algebraic structure of the corpus, from which novel equations are sampled. To ensure the generated inputs lie within a physically meaningful domain, the applicability domain is characterised for each equation through non-intrusive probing, also recovering inter-variable constraints. Input sampling further mimics realistic experimental conditions by drawing from random sub-ranges of the valid domain with mixed uniform and truncated normal distributions. The generated data is statistically validated against the Feynman equation corpus using Kolmogorov-Smirnov tests. The generated equations match the corpus on all of the eight studied structural features, compared to only two for an unsmoothed purely probabilistic grammar, demonstrating that the Bayesian prior is essential for structural fidelity given the size of the corpus. In a downstream hyperparameter-tuning task, a gradient-boosted regressor tuned on the synthetic data picks, on average, the 6th-best configuration out of 20 on real data, matching the result of tuning on real data itself and substantially outperforming random expression trees (10th) and noise (19th).
arXiv:2606.06959v1 Announce Type: new
Abstract: Hallucination detection is essential for the reliable deployment of large language models (LLMs). However, existing evaluations face two core challenges: inconsistent inference configuration and evaluation, and limited coverage of downstream domains and tasks. Consequently, reported detector performance is often difficult to compare, reproduce, and generalize beyond specific experimental settings. We introduce OpenHalDet, a unified benchmark for hallucination detection across diverse generation scenarios. OpenHalDet standardizes the evaluation pipeline, from prompt construction and response generation to truthfulness annotation, detector scoring, and metric computation. It supports heterogeneous detector families under different access settings, including black-box methods that use only generated outputs, gray-box methods that rely on probability-based signals, and white-box methods that exploit internal model signals. By bringing diverse tasks, models, and detectors into a shared framework, OpenHalDet enables controlled comparison and provides a systematic view of how different detection paradigms behave in LLM applications. We release OpenHalDet as an open and extensible codebase to facilitate reproducible evaluation and future development of hallucination detection methods. The code and datasets are available at https://github.com/Nellie179/Hallucination-Detection.
arXiv:2606.06960v1 Announce Type: new
Abstract: Experience-based self-evolution is crucial for LLM agents, but existing benchmarks often assume explicit goals, stable task patterns, and clear feedback. We study a more challenging setting: low-repetition tasks with implicit rewards, where past experience is difficult to reuse and feedback is delayed, noisy, and outcome-level. We introduce \textsc{FinEvolveBench}, a temporally controlled benchmark for financial sentiment prediction that links daily news-driven predictions to future excess returns. We further propose Tree-of-Experience (ToE), a structured experience-management method that organizes, retrieves, validates, and updates agent experience. Experiments show that general-purpose experience mechanisms do not consistently outperform no-experience baselines, while ToE achieves stronger overall performance. These results highlight the importance of structured experience management for self-evolving agents in implicit-reward environments.