Modern video models possess strong visual priors, yet do not natively represent human action in metric 3D space. We present EgoGeneralist, which adapts pretrained video generators to unify action-conditioned egocentric video generation, pose estimation, and joint action-video generation. Our role-adaptive action renderer provides video-native representations for multidirectional action-video learning, making metric motion usable as both a spatial control signal and a recoverable generation target. Cross-stream action-video coupling establishes close correspondence between metric motion and visual dynamics through lightweight adaptation while preserving the pretrained visual prior. Together, these designs establish a generalist framework spanning five action-video tasks. Despite adaptation only on controlled laboratory recordings, EgoGeneralist generalizes to diverse unseen scenes and interactions with effective action control and visual quality, bridging pretrained visual knowledge and metric 3D human motion.
Molecular representations such as fingerprints and learned embeddings play a central role in cheminformatics and small-molecule machine learning. They serve as the primary input for a diverse array of tasks, including clustering, dataset splitting, and virtual screening, yet their evaluation is often limited to downstream model performance on narrow benchmark tasks. We carried out a systematic investigation of the quality of molecular representations, finding that none of the state-of-the-art molecular embedding models capture graph structure, 3D shape, or ADMET endpoint similarity better than fingerprints or physicochemical descriptors. However, classical fingerprints also revealed their own material limitations. For example, we show that depending on the representation, between 225M and $\sim$4B molecules have non-unique representations in the 10.8B Enamine REAL database, with implications for virtual screening. Our findings provide an insight into the current limitations of molecular representations, outlining future research directions for the development of more robust encodings of small molecules.
In many domains, practitioners seek models that produce accurate forecasts while faithfully capturing latent system dynamics. Existing approaches typically sacrifice one of these goals: deep state space models often assume Gaussian latent transitions, limiting fit and forecasting, while diffusion models are highly expressive but lack principled inference for the underlying dynamics. To combine the strengths of both, we introduce the Diffusion-Driven State Space Model (DDSSM), which replaces the conventional Gaussian transition distribution with a diffusion model. Our DDSSM resolves the open problem of how to jointly train an autoencoder and a diffusion model on sequential data, thereby extending the literature on latent diffusion models for time series. Moreover, we find that the DDSSM empirically outperforms a state-of-the-art deep SSM at fitting and forecasting a simulated time series with multimodal transitions.
Connecting hand input in mixed reality (MR) to a nearby physical device typically requires an integration written in advance for that device, its transport and the mapping. ControlAnything defers this integration to the moment of use and assigns it to a coding agent. On a Meta Quest 3, the wearer selects an object outlined by SAM 3.1 and states what their hands should control. An agent built on Meta's Muse Spark model then writes a connector program on the wearer's laptop, tests it against the live device through hand-written MIDI, Bluetooth and key-press transports, and returns augmented-reality (AR) controls, a per-frame hand program, or both. These outputs are a form of generative UI, interfaces created by an AI model; here, the agent also generates and tests the code that binds the interface to a physical device. In a single-session technical evaluation with the author as the only wearer, nine of ten requests addressing five targets reached a ready state, and the wearer stopped one. The median time to ready was 32.5 s (range 24.9-57.8 s, excluding speech transcription); sequential model calls accounted for a median of 89.9% of this time, and 91.7% of input tokens came from the prompt cache. The system's evidence of success extends to transport acknowledgements but not to physical effects. These findings point to connector reuse and the observation of effects through the headset's own sensors as next steps for on-demand device integration in MR. A user study with first-time wearers is planned.
Automatic sleep stage classification from electroencephalography (EEG) signals is a critical task in clinical sleep medicine. While deep learning models have shown promising results in this domain, their performance is fundamentally constrained by the quality of training labels, which human experts manually annotate and that are subject to inter-rater disagreement and systematic labeling errors. In this paper, we propose a novel LLM-assisted label-noise filtering pipeline that operates exclusively during training to identify and remove unreliable sleep-stage annotations before retraining the classifier. Specifically, TinySleepNet is first trained on the original dataset to extract per-epoch SoftMax confidence scores, then low-confidence epochs are submitted to LLaMA 3.1 along with their EEG spectral features and surrounding sequence context. The LLM evaluates each epoch against AASM sleep-scoring guidelines and assigns a verdict of RELIABLE or SUSPICIOUS. Suspicious epochs are removed from the training set, and TinySleepNet is retrained on the cleaned data. Evaluated on the Sleep-EDF Expanded dataset using 20-fold cross-validation, our method improves overall accuracy from 81.8% to 84.6% and Macro F1-score from 75.6% to 77.4%, with particularly notable gains in REM detection. Crucially, because the LLM operates only during training, the lightweight TinySleepNet architecture (1.3M parameters) is preserved at inference, making the proposed pipeline suitable for real-time, wearable sleep monitoring applications.
Mathis Immertreu, Fitim Abdullahu, Thomas Kinfe, Claus Metzner, Helmut Grabner, Patrick Krauss, Achim Schilling
Multimodal language models can simulate different observers, yet it remains unclear how assigned perspectives reshape their internal representations and whether these changes govern their judgements. Here, we investigate the representational geometry of persona-conditioned visual interestingness in Qwen3-VL-8B-Instruct. Across 100 images and 53 prompting conditions, we systematically vary persona content and the linguistic framing used to introduce it, while recording behavioural ratings and layer-wise hidden states. We observe a systematic reorganisation of representational geometry across processing depth: early layers primarily encode image identity, whereas persona-specific structures emerge in later layers. Strikingly, prompt framing reshapes internal representations more strongly than persona identity, whereas persona content has a greater impact on behavioural judgements. This dissociation is further illustrated by minimal identity statements that alter representations even without changing responses. These findings demonstrate that representational distance does not necessarily reflect behavioural relevance. More broadly, they reveal how simulated perspectives emerge as distinct structures within the representational space of artificial cognitive systems, highlighting a fundamental distinction between the geometry of internal representations and their functional consequences.
Large language model agents increasingly retrieve tools from large catalogs, yet most retrievers rank each tool independently by request similarity. For multi tool tasks, this recovers the goal matching tool while omitting prerequisite tools required by the workflow. We introduce LIBRA, a framework for Library Induction by Runtime Assembly that turns a flat tool catalog into an executable tool library. An LLM proposes per tool contracts of candidate invocations, reusable outputs and state changes, and success conditions; LIBRA then tests these compositions in a sandbox and retains the relations supported at runtime. Constructed before user queries, the library requires no dependency annotations, tool use trajectories, or task specific training, and supports both one-shot and agentic retrieval. On AppWorld, BFCL, and SciToolEval, LIBRA improves one-shot Recall and NDCG over the same dense retriever by 7.9 and 3.7 percentage points on average. Within ToolReAGt, it yields corresponding gains of 4.5 and 2.9 points after ten retrieval iterations.
Multi-modal large language models (MLLMs) with impressive perception and understanding capabilities will enable a range of downstream applications. However, their substantial computational and memory requirements hinder the real-world implementation, particularly for resource-constrained edge devices. Quantization has emerged as an effective solution for deploying large models. While post-training quantization (PTQ) successfully retains the performance at moderate bit-width (e.g., W8A8), it suffers from severe performance degradation under low-bit quantization (e.g., W4A8), especially for small-scale models that are more suitable for edge deployment. In this paper, we propose ProQuant, a novel weight-activation quantization-aware training (QAT) framework tailored for low-bit MLLMs. First, to align the limited finetuning dataset distribution with the pretrained dataset distribution, we propose a self-relabeling (SRL) scheme to resample the responses of the open-source multimodal dataset. Second, we are the first to take a decoupled view of multi-modal (MM) understanding and language modeling abilities. We propose a progressive block activation (PBA) mechanism to prioritize the MM understanding recovery. We also introduce a vision-sensitive factor $\vecτ$ that effectively identifies the MM-critical layers. Extensive experiments across various benchmarks, covering models with 2B-8B parameters, show that ProQuant significantly outperforms existing methods. For example, our W4A8 Qwen3-VL-2B-Instruct improves average accuracy by impressively 6.6% and achieves performance comparable to full precision counterparts with only approximately 1% performance drop. Code is available at https://github.com/yufeixue-ai/ProQuant/.
Hamid Sarmadi, Ola Hall, Thorsteinn Rögnvaldsson, Mattias Ohlsson
This paper investigates the novel application of Large Language Models (LLMs) with vision capabilities to analyze satellite imagery for village-level poverty prediction. Although LLMs were originally designed for natural language understanding, their adaptability to perform multimodal tasks, including geospatial analysis, has opened new frontiers in data-driven research. By leveraging advancements in vision-enabled LLMs, we assess their ability to provide interpretable, scalable, and reliable insights into human poverty from satellite images. Using a pairwise comparison approach, we demonstrate that ChatGPT can rank satellite images based on poverty levels with an accuracy comparable to a model trained on human-expert-identified image features and better than direct expert wealth classifications reported previously. These findings highlight both the promise and the limitations of LLMs in socioeconomic research, providing a foundation for their integration into poverty assessment workflows. This study contributes to the ongoing exploration of unconventional data sources for welfare analysis and opens pathways for cost-effective, large-scale poverty monitoring. Our results also put forward the question of how reliable public datasets, such as DHS, are for retrieving wealth indices that are anonymized by adding noise to the location coordinates. The code and data used for the analyses in this paper are publicly available.
The Rectified Power Unit (RePU) activation function, a differentiable generalization of the Rectified Linear Unit (ReLU), has shown promise in constructing neural networks due to its smoothness properties. However, deep RePU networks often suffer from critical issues such as vanishing or exploding values during training, rendering them unstable regardless of hyperparameter initialization. Leveraging the perspective of effective field theory, we identify the root causes of these failures and propose the Modified Rectified Power Unit (MRePU) activation function. MRePU addresses RePU's limitations while preserving its advantages, such as differentiability and universal approximation properties. Theoretical analysis demonstrates that MRePU satisfies the criticality conditions in the limit of a small kernel, and a finite-kernel analysis yields an initialization rule for deep MRePU networks. Experiments on polynomial regression, physics-informed neural networks (PINNs) and real-world vision tasks show that MRePU networks initialized by this rule remain trainable at depths at which RePU networks fail, while their accuracy relative to standard activations depends on the task.
Mohammad Rowhani Sistani, Katarzyna Kosek-Szott, Pierluigi Gallo
Wireless links deployed in orchards often exhibit significant variability in the strength of the received signal that is not adequately captured by classical distance-based propagation models. In row-structured olive groves, signal attenuation differs markedly between along-row and cross-row propagation directions, leading to discrepancies when using omnidirectional propagation assumptions such as those adopted in the Free Space Path Loss (FSPL) model or ITU-R vegetation loss formulations. This paper proposes a topology-based propagation model that explicitly accounts for orchard layout and the relative positions of radio devices within the plantation structure. Experimental validation was conducted using LoRa technology operating at 868 MHz, and the results were compared with established models from the literature and with the proposed two-dimensional model. The proposed approach achieves a closer fit to measured RSSI data than conventional models, providing a more reliable basis for link budgeting and network planning in structured agricultural environments.
Brain foundation models have transformed medical artificial intelligence and computational neuroscience by reducing the reliance on task-specific data through large-scale pretraining. While they have been successful with non-invasive neuroimaging data, their adoption in functional near-infrared spectroscopy (fNIRS) has lagged due to the limited size and heterogeneity of public datasets, which are collected under varying protocols, devices, and sensor configurations. In addition, no single fNIRS dataset is comparable in size to the largest datasets in functional magnetic resonance imaging (fMRI), electroencephalography (EEG), or magnetoencephalography (MEG). The fNIRS datasets are also spread across different platforms and often have insufficient information, or lack the protocols that would make them easier to use. Given the advances in EEG foundation models that enable flexible modeling across diverse EEG montages, we curated 1,955 hours of fNIRS data from 67 public datasets spanning 2,798 participants, representing the largest fNIRS curation effort to date. The curated data include oxygenated (HbO) and deoxygenated (HbR) hemoglobin signals along with the three-dimensional position of every channel in the standard space of the Montreal Neurological Institute (MNI), which makes positions comparable across datasets. We release dataloaders that handle each dataset automatically, along with the minimal preprocessing pipeline used to prepare the data, at https://github.com/sposso/fnirs-datasets. This work aims to facilitate the future development of foundation models for fNIRS.
Mateo Clémente, George Z. Xue, Amin Abyaneh, Liheng Ma, Rui Heng Yang, Tongtong Cao, Zhanguang Zhang, Yingxue Zhang
World-action models adapt pretrained video generators, and vision-language-action models build on pretrained vision-language backbones; both tie policy architecture and inference cost to large transformers designed for generation rather than control. It remains unclear how much these generation-oriented choices affect control. In this work, we introduce Latent Action JEPA (La-JEPA), a compact world-action model whose policy transformer is trained from scratch in the feature space of a pretrained V-JEPA 2.1 encoder. Following Fast-WAM, La-JEPA learns to predict future features jointly with actions during training and generates only actions at inference. Among policies without embodied pretraining, La-JEPA achieves state-of-the-art performance in clean and randomized RoboTwin 2.0 scenes and on LIBERO-Plus with correct instructions, while using far fewer parameters than the strongest competing policies. It runs substantially faster than Fast-WAM at comparable performance and outperforms the closest JEPA-based methods, LeapBot-WA and JEPA-WAM, on LIBERO-Plus and RoboTwin 2.0 without their pretrained policy backbones. On four real-robot tasks, its average success is comparable to pi0.5 and above non-embodied-pretrained baselines. Through a design study of the choices behind La-JEPA, we show that JEPA features outperform video VAE latents, most clearly out of distribution, and that training-time future prediction choices significantly impact the policy even when future prediction is skipped at inference. We correct the LIBERO-Plus instructions, as the officially released evaluator corrupts them for most variants, and re-evaluate several strong open baselines. Without a generative pretrained policy backbone, La-JEPA is better and faster than far larger policies, suggesting that robot foundation models may benefit from architectures designed for control rather than inherited from generation.
Linear attention has recently gained significant attention for long-context inference due to its constant decoding cost with respect to context length. However, existing serving systems typically serve linear attention by recurrently computing and updating a large linear attention state in every decoding step. Since the state is much larger than the per-token key and value, recurrent decoding incurs substantial memory access and becomes inefficient for serving linear attention. In this paper, we propose KVBuffer, an IO-aware serving mechanism for linear attention. By buffering recent keys and values, KVBuffer enables serving systems to compute linear attention outputs in more flexible and memory-efficient ways. For decoding, KVBuffer enables chunkwise computation, which reduces average memory access and decoding latency by deferring state updates and applying them in batch. For speculative decoding, KVBuffer verifies draft tokens in parallel and avoids storing temporary states. For short contexts, KVBuffer computes attention outputs directly from buffered keys and values, without creating or updating the linear attention state. We implement KVBuffer in SGLang for Qwen3.5 and Nemotron-3 series models. Our evaluations show that KVBuffer can speedup batch decoding by up to 1.47x and increase the maximum number of serving requests by 5x for speculative decoding when verifying four draft tokens.
Jaime P. Pérez-Moro, Alejandro Romero, Hector Quintian, Francisco Bellas
Robot Operating System (ROS) is becoming a standard in robotics courses at higher education, but its installation, configuration and distributed programming model can introduce a significant entry barrier in the new landscape of AI-based master and degree studies, where many students do not have a strong informatics background. This paper presents an educational experience that enables students to program a TurtleBot3 Burger directly in Python, while preserving a direct transfer path from simulation to the physical robot. The approach is based on a custom library that mirrors the Webots API, allowing the same controller code to be executed in the simulator and on the real TurtleBot3 with minimal modification. The library provides access to the robot motors, wheel encoders, inertial sensors and LiDAR, and is complemented by scripts that automate deployment and execution on the physical platform. It was used in a master-level course on introductory autonomous vehicles, and initial student feedback suggests that the approach reduces the perceived complexity of ROS-related infrastructure and helps learners focus on algorithmic logic and real-world behaviour.
Despite recent progress in handwritten text recognition, medieval Arabic scientific manuscripts remain an open challenge for current vision-language models (VLMs) due to cursive ligatures, variable diacritization, and more importantly the lack of training data. This paper aims to address this gap, by proposing the first annotation guidelines dedicated to the transcription of scientific Arabic manuscripts inspired by textual criticism, a well-established approach to historical text editing. We then propose, ATHAAR, a new dataset manually annotated following this guideline. Unlike previous resources in the field, ATHAAR varies both witnesses of the same text, calligraphic styles, and content to deal with texts but also numerals. Relying on state-of-the-art VLMs, we finally design a set of experiments to study models transfer across scripts, chapters, and treatises. Our findings highlight the limitations of current systems when faced with the specificities of ancient scientific texts.
Automated fact-checking tools, cheap and fast but imperfect, are increasingly available on social platforms. We study when such tools can backfire in a sequential social learning model of misinformation. Agents observe an article's reliability and their predecessors' actions, which indicate whether they liked or disliked the article. They may also pay to fact-check it. We examine how welfare, defined as the probability of a correct cascade, changes when a free, error-prone automated tool supplements costly but perfectly accurate manual fact-checking. Two opposing effects arise. Broader use of fact-checking helps agents escape wrong cascades. At the same time, errors from the automated tool can mislead agents who would otherwise have verified the article perfectly, reducing the likelihood of correct cascades. We prove that adding the tool strictly lowers welfare for articles of intermediate reliability when truthful articles are not too informative and manual checking is sufficiently costly. Simulations show that the tool can help on either side of the intermediate range, and that it improves on manual-only checking on average only above a threshold accuracy.
This paper advances the theory of reproducing kernel Banach spaces (RKBSs) by investigating the general $\ell^p$ regularization framework, with a particular focus on the non-convex regime $0<p<1$. While our preceding work established the foundation for RKBSs with the $\ell^1$ norm, the pursuit of stronger sparsity---closer to the $\ell^0$ pseudo-norm---naturally motivates this exploration into $\ell^p$ quasi-norms. For $0<p<1$, we construct reproducing kernel quasi-Banach spaces (RKQBSs) with $\ell^p$ quasi-norms and establish existence and sparsity results for interpolation and regularized learning under appropriate assumptions. We characterize the linear and relaxed representer theorems in terms of a generalized Lebesgue constant and derive bounds under kernel decay, sample spacing, and stability assumptions. For $p>1$, we construct RKBSs using a fixed countable dictionary, establish minimizer existence, and identify the optimal relaxed representer factor through an off-sample coefficient operator. These results reveal fundamental differences between the regimes $0<p<1$, $p=1$, and $p>1$. Numerical experiments compare the accuracy and sparsity of solutions for these three cases.
While teaching untyped $λ$-calculus to undergraduate students, we were left wondering why $α$-equivalence is not directly inductively defined. In this paper, we demonstrate that this is indeed feasible. Specifically, we provide a grounded, inductive definition for $α$-equivalence and show that it conforms to the specification provided in the literature. The work presented in this paper is fully formalized in the Rocq Prover.
Renjie Zhao, Yuxiang Wu, Mingyu Zhang, Jiaxin Li, Sisi Li, He Li, Yimin Sheng, Tianxi Tan 等 12 位作者
World models have recently achieved impressive progress in visual prediction and interactive generation, but extending them to multi-agent environments introduces a fundamental scalability challenge. Existing methods generally assume a fixed number of agents during training and inference, which ties the model to a pre-determined agent population and limits inference-time scalability. Our key insight is that cross-view consistency should arise from a shared world state whose evolution does not assume a predefined number of agents, while agent-specific observations should be generated by querying this state through a unified rendering interface. Based on this insight, we propose Khora, a scalable multi-agent world model that supports inference-time expansion to arbitrary numbers of agents without retraining. Our framework decouples world-state evolution from visual rendering and introduces a population-agnostic rendering mechanism for incorporating other agent information. This design maintains cross-view consistency through the shared world state rather than through dense interactions among observation streams inside the expensive video generator, enabling approximately linear practical scaling with the number of queried views. Qualitative experiments demonstrate that our approach generalizes to unseen numbers of agents while maintaining visual quality and multi-agent consistency. We further implement a real-time interactive system to demonstrate scalable open-world simulation.
Spiking Neural Networks (SNNs) are promising for energy-efficient, real-time edge computing, yet their performance is often constrained by the limited adaptability of conventional leaky integrate-and-fire (LIF) neurons. Existing LIF models struggle with restricted information capacity and susceptibility to noise, leading to degraded accuracy and compromised robustness. Inspired by the dynamic self-regulation of biological potassium channels, we propose the Potassium-regulated LIF (KvLIF) neuron model. KvLIF introduces an auxiliary conductance state that integrates membrane potential and spiking history to adaptively modulate neuronal excitability and reset dynamics. This design extends the dynamic response range of neurons to varying input intensities and effectively suppresses noise-induced spikes. We extensively evaluate KvLIF on both static image and neuromorphic datasets, demonstrating consistent improvements in classification accuracy and superior robustness compared to existing LIF models. Our work bridges biological plausibility with computational efficiency, offering a neuron model that enhances SNN performance while maintaining suitability for low-power neuromorphic deployment.
Multimodal large language models (MLLMs) are widely applied to visual document understanding. However, comprehending long documents remains an issue by the limited context window. Though recent multimodal retrieval-augmented generation (MMRAG) can address this challenge by retrieving relevant pages. It still struggles with the visual question answering (VQA) requiring holistic comprehension of a document. To cope with this, knowledge graph (KG) that summarizes global knowledge of a document can provide an effective solution. However, most existing LLM-based KG construction methods handle only the language modality, leaving the automatic creation of multimodal KGs (MMKGs) for visually rich documents largely unexplored. In this paper, we introduce a multimodal graph-based RAG approach to tackle this problem. Existing LLM-based KG methods evaluate the QA performance relying on indirect evidence such as comprehensiveness, diversity, empowerment, and so on. The lack of annotated datasets for comprehensive document-level VQA poses a significant challenge to effective model evaluation. To overcome this limitation, we also introduce a new benchmark, DLVQA (document-level VQA), which provides reference summaries and corresponding supporting facts for global document-level questions. Experimental results show that our approach outperforms existing MMRAG or KG-based approaches on multi-hop QA/VQA benchmarks and DLVQA.
Francisco Pereira, Justine Sherry, Pratiksha Thaker
In 2003, Clark et al. proposed the "Knowledge Plane", a layer of intelligent agents that share rich information about network state in order to collaboratively manage the Internet across administrative boundaries. The technology of the day lacked the two capabilities it demanded most: reasoning and unstructured communication. Thus, automated network management has remained largely confined to single domains, while interdomain problems are still resolved by human operators on the phone. We argue that agentic large language models can finally unlock the potential of the Knowledge Plane, but that doing so will require solving new research problems that arise particularly in this decentralized, multi-agent, coopetitive setting that have received limited attention in the literature thus far.
Linear Complementarity Problems (LCPs) with sufficient matrices form a well-studied class of problems, dating back from the foundational works of Cottle and Dantzig (1968). LCPs capture convex quadratic programs, and admit reductions from problems such as discounted stochastic games or mean-payoff games. While the original approach for solving these problems was based on pivoting methods, in their 1991 book, Kojima, Megiddo, Noma, and Yoshise introduced an interior point method (IPM) with running time that is polynomial in the input size, but which depends linearly on a certain parameter called the handicap number of the matrix. Despite the ensuing rich literature on sufficient LCP IPMs, the linear dependence on the handicap number has not been improved since then. We present an IPM whose running time depends on the square root of the handicap number. The algorithm can decrease the duality gap by a constant factor in $O\left(\min\left\{\sqrt{n} (1+κ),n^{9/8}\sqrt{1+κ}\right\}\right)$ iterations. Curiously, this can already be achieved using standard Newton updates. We analyze the IPM in a norm defined by a local weighting, where centralizing Newton steps quadratically converge to the central path, without any dependence on the handicap number. The weights arise from a characterization by Rump (Linear Algebra and Its Applications, 1997, 1998, 2003) of the sign-real spectral radius and its relation to sufficiency; another ingredient of our proof is a new operator norm characterization of the handicap number. Moreover, the connection to Rump's work also enables us to show coNP-completeness of testing sufficiency, and approximation hardness of the handicap number.
Daniel Collin, Yuri Shprits, Luca Chiarabini, Stefan J. Hofmeister, Nadja Klein, Guillermo Gallego
Accurately predicting fast solar wind conditions is challenging, as uncertainties are large and unquantified by traditional single-value prediction models. In particular, the risks of high-speed solar wind streams (HSSs), which can cause damage to technological infrastructure, cannot be reliably assessed without probabilistic forecasts. We present PROSWIN, a probabilistic machine learning model that forecasts the hourly solar wind speed (SWS) at Earth with a four-day lead time. The approach combines solar images and magnetograms using a deep neural network coupled to a distributional regression algorithm. Because standard error metrics underweight the relevance of HSS peaks, we further introduce the prediction score, a model-selection metric that jointly rewards timeline and HSS peak accuracy. On 14 years of data, our forecast achieves very well-calibrated uncertainties (<1% average deviation). Using the continuous ranked probability score (CRPS), a metric that assesses distributional accuracy, we obtain a timeline CRPS of 41.0 km/s, an HSS peak CRPS of 45.3 km/s, and a prediction score of 42.3 km/s. We find that the 171 Å channel is an important complement to the typically used 193 Å and 211 Å channels and that the prediction score for model selection improves the applicability of the model. Compared to selected models from the literature, ours is the only one that is accurate for both timeline and HSS peak values, rather than trading one off against the other. These results support the advantages of probabilistic over single-value solar wind models. The introduced methods are also transferable to other forecasting problems.
Financial document recognition is often assessed using character and word error rates, although these metrics do not express whether a system preserves the financial facts on which people and organisations act. This paper develops a human-centred, risk-constrained framework for evaluating financial document recognition as evidence recovery and decision support. It defines typed financial facts linked to visible source regions and separates verbatim transcription, deterministic normalization, and business interpretation. A layered quality model covers text, localization, layout, tables, fields, fact correctness, evidence grounding, uncertainty, robustness, and operational performance. Non-compensatory acceptance gates use one-sided uncertainty bounds for critical-field errors, preventing strong aggregate scores from masking failures in amounts, identifiers, signs, periods, or entity associations. Selective automation is formulated as a constrained routing problem that accounts for residual risk, human-review capacity, and review cost. The framework further specifies annotation and adjudication procedures, leakage-resistant benchmark partitions, statistical estimands, subgroup analyses, and a reproducibility card. Formal results establish the limitations of compensatory scoring, the non-zero upper risk bound after zero observed errors, and a loss sensitive decision rule for routing uncertain cases to human review. The contribution is theoretical and methodological; no new model, dataset, or empirical performance result is claimed.
Orhun Görkem, Athanasios Vasilopoulos, Georgios Pagonis, Victoria Clerico, Hadjer Benmeziane, Irem Boybat, Hsinyu Tsai, Abu Sebastian 等 9 位作者
Large Language Models (LLMs) incur severe memory-access overheads due to the cost of loading large weight matrices and key-value (KV) caches. Analog In-Memory Computing (AIMC) is an emerging solution that enables computation directly on static weights stored in memory, substantially reducing energy and latency. 3D-IMC further stacks memory arrays into vertical tiers, increasing capacity but revealing a new challenge: how weights are placed in memory determines which matrix-vector multiplications (MVMs) can execute concurrently, affecting latency and throughput. The concurrency aspect of 3D-IMC weight-mapping remains underexplored in the literature, leading to sub-optimal utilization of compute resources on 3D-IMC chips. In this context we propose CAMP (Context-Aware Mapping and Placement), an inference-context-aware weight mapping algorithm for 3D-IMC-based LLM accelerators. CAMP captures the potential resource contention resulting from the architecture, targeted network, and runtime context in a conflict matrix, enabling throughput-optimized weight mapping. This mapping is refined through a profiling-driven feedback loop that efficiently prunes the search space and returns optimized mappings within practical runtimes. CAMP is an offline, software-based solution requiring no hardware changes and improving 3D-IMC compute utilization from 42.4% to 60.3%, delivering up to 2.0x higher 3D-IMC throughput and 1.74x higher end-to-end throughput over state-of-the-art mapping algorithms.
Enhan Zhao, Yongqi Tong, Xuejun Yu, Hang Wang, Pan Wang, Yichao Wang, Di He, Jianshe Li 等 10 位作者
Existing long-horizon agent benchmarks often emphasize final task success while providing limited assessment of intermediate execution validity. Yet real-world business operations require agents to navigate extensive, potentially outdated or conflicting policies while satisfying critical constraints throughout execution. These requirements are particularly consequential in high-stakes domains such as payments, where intermediate violations can incur financial losses or regulatory consequences regardless of the final outcome. To bridge this evaluation gap, we introduce BizAgentBench, a multi-turn benchmark grounded in real-world procedures from a leading payments company. Its executable environment integrates 789 knowledge documents and 601 discoverable business tools, with consistent policies, states, and tool behavior. The benchmark comprises 50 tasks involving resource constraints, temporal dependencies, prior commitments, and conflicting or outdated guidance. Each task is validated through reference execution, while deterministic checks assess both goal attainment and execution validity without prescribing a single solution path. Experiments reveal substantial limitations even among frontier models: Claude Opus 5 and GPT-6 Sol achieve strict success rates of only 64.0% and 61.3%, respectively. For Qwen3.8-27B, goal attainment reaches 42.0%, but strict success drops to 32.7%, demonstrating how outcome-based evaluation can mask invalid execution. Beyond success rates, DeepSeek-V4.1-Flash averages 191.3 tool calls per run, approximately 70% more than reference executions, revealing substantial execution overhead. Overall, BizAgentBench establishes a testbed for evaluating and advancing long-horizon agents beyond task completion, toward reliable and constraint-compliant execution in complex business environments.
Online conformal prediction (OCP) seeks prediction intervals that achieve long-run $1-α$ coverage for arbitrary (possibly adversarial) data streams, while remaining as informative as possible. Existing OCP methods often require manual learning-rate tuning to work well, and may also require algorithm-specific analyses. Here, we develop a general regret-to-coverage theory for interval-valued OCP based on the $(1-α)$-pinball loss. Our first contribution is to identify \emph{linearized regret} as a key notion, showing that controlling it implies coverage bounds for any online algorithm. This relies on a black-box reduction that depends only on the Fenchel conjugate of an upper bound on the linearized regret. Building on this theory, we propose UP-OCP, a parameter-free method for OCP, via a reduction to a two-asset portfolio selection problem, leveraging universal portfolio algorithms. We show strong finite-time bounds on the miscoverage of UP-OCP, even for polynomially growing predictions. Extensive experiments support that UP-OCP delivers consistently better size/coverage trade-offs than prior online conformal baselines.
Expert pruning compresses Mixture-of-Experts (MoE) language models by removing the experts that a calibration set marks as unimportant. Studies of high-resource languages largely find that English calibration transfers well to other languages. We test this for Lithuanian: we pruned four MoE models with REAP, removing 25% or 50% of the experts with English, Lithuanian or mixed calibration, and evaluated both languages on reading comprehension, translation and language modelling. English calibration does not transfer to Lithuanian: at 25% pruning, English comprehension fell by a median of 1.7 points but Lithuanian by 21.4. The transfer is asymmetric. Lithuanian calibration removed most of the Lithuanian loss and never significantly hurt English comprehension, which was on average 2.5 points better than under English calibration; only English language modeling got slightly worse. Routing suggests why: in the last layers, Lithuanian sends 55-72% of its expert selections to just eight experts, against 19-39% for English, and English calibration removes some of these experts. Compressed models should be calibrated, and evaluated, in every language they must serve.