Human activity is moderated by norms. However, machines are often trained without explicit supervision on norm understanding and reasoning, particularly when norms are physically- or socially-grounded. To improve and evaluate the normative reasoning capability of vision-language models (VLMs), we present \dataset{} $\|\epsilon\|$, consisting of 1,853 challenging, multi-stage MCQ questions based on ego-centric videos of human interactions, evaluating both the prediction and justification of normative actions. The normative actions encompass seven categories: safety, privacy, proxemics, politeness, cooperation, coordination/proactivity, and communication/legibility. To compile this dataset at scale, we propose a novel pipeline leveraging video sampling, automatic answer generation, filtering, and human validation. Our work demonstrates that current state-of-the-art vision-language models lack robust norm understanding, scoring a maximum of 54\% on \dataset{} (versus a human bench of 92\%). Our analysis of performance in each dimension highlights the significant risks of safety, privacy, and the lack of collaboration and communication capability when applied to real-world agents. We additionally show that through a retrieval-based generation (RAG) method, it is possible to use \dataset{} to enhance normative reasoning in VLMs.
Formal mathematical reasoning remains a critical challenge for artificial intelligence, hindered by limitations of existing benchmarks in scope and scale. To address this, we present FormalMATH, a large-scale Lean4 benchmark comprising 5,560 formally verified problems spanning from high-school Olympiad challenges to undergraduate-level theorems across diverse domains (e.g., algebra, applied mathematics, calculus, number theory, and discrete mathematics). To mitigate the inefficiency of manual formalization, we introduce a novel human-in-the-loop autoformalization pipeline that integrates: (1) specialized large language models (LLMs) for statement autoformalization, (2) multi-LLM semantic verification, and (3) negation-based disproof filtering strategies using off-the-shelf LLM-based provers. This approach reduces expert annotation costs by retaining 72.09% of statements before manual verification while ensuring fidelity to the original natural-language problems. Our evaluation of state-of-the-art LLM-based theorem provers reveals significant limitations: even the strongest models achieve only 16.46% success rate under practical sampling budgets, exhibiting pronounced domain bias (e.g., excelling in algebra but failing in calculus) and over-reliance on simplified automation tactics. Notably, we identify a counterintuitive inverse relationship between natural-language solution guidance and proof success in chain-of-thought reasoning scenarios, suggesting that human-written informal reasoning introduces noise rather than clarity in the formal reasoning settings. We believe that FormalMATH provides a robust benchmark for benchmarking formal mathematical reasoning.
Foundation models constitute a significant advancement in computer vision: after a single, albeit costly, training phase, they can address a wide array of tasks. In the field of Earth observation, over 75 remote sensing vision foundation models have been developed in the past four years. However, none has consistently outperformed the others across all available downstream tasks. To facilitate their comparison, we propose a cost-effective method for predicting a model's performance on multiple downstream tasks without the need for fine-tuning on each one. This method is based on what we call "capabilities encoding." The utility of this novel approach is twofold: we demonstrate its potential to simplify the selection of a foundation model for a given new task, and we employ it to offer a fresh perspective on the existing literature, suggesting avenues for future research. Codes are available at https://github.com/pierreadorni/capabilities-encoding.
Neurosymbolic approaches integrating large language models with formal reasoning have recently achieved human-level performance on mathematics competition problems in algebra, geometry and number theory. In comparison, combinatorics remains a challenging domain, characterized by a lack of appropriate benchmarks and theorem libraries. To address this gap, we introduce CombiBench, a comprehensive benchmark comprising 100 combinatorial problems, each formalized in Lean~4 and paired with its corresponding informal statement. The problem set covers a wide spectrum of difficulty levels, ranging from middle school to IMO and university level, and span over ten combinatorial topics. CombiBench is suitable for testing IMO solving capabilities since it includes all IMO combinatorial problems since 2000 (except IMO 2004 P3 as its statement contain an images). Furthermore, we provide a comprehensive and standardized evaluation framework, dubbed Fine-Eval (for $\textbf{F}$ill-in-the-blank $\textbf{in}$ L$\textbf{e}$an Evaluation), for formal mathematics. It accommodates not only proof-based problems but also, for the first time, the evaluation of fill-in-the-blank questions. Using Fine-Eval as the evaluation method and Kimina Lean Server as the backend, we benchmark several LLMs on CombiBench and observe that their capabilities for formally solving combinatorial problems remain limited. Among all models tested (none of which has been trained for this particular task), Kimina-Prover attains the best results, solving 7 problems (out of 100) under both ``with solution'' and ``without solution'' scenarios. We open source the benchmark dataset alongside with the code of the proposed evaluation method at https://github.com/MoonshotAI/CombiBench/.
Large Language Models (LLMs) have demonstrated significant promise for various applications in healthcare. However, their efficacy in the Arabic medical domain remains unexplored due to the lack of high-quality domain-specific datasets and benchmarks. This study introduces MedArabiQ, a novel benchmark dataset consisting of seven Arabic medical tasks, covering multiple specialties and including multiple choice questions, fill-in-the-blank, and patient-doctor question answering. We first constructed the dataset using past medical exams and publicly available datasets. We then introduced different modifications to evaluate various LLM capabilities, including bias mitigation. We conducted an extensive evaluation with five state-of-the-art open-source and proprietary LLMs, including GPT-4o, Claude 3.5-Sonnet, and Gemini 1.5. Our findings highlight the need for the creation of new high-quality benchmarks that span different languages to ensure fair deployment and scalability of LLMs in healthcare. By establishing this benchmark and releasing the dataset, we provide a foundation for future research aimed at evaluating and enhancing the multilingual capabilities of LLMs for the equitable use of generative AI in healthcare.
The Limit Order Book (LOB), the mostly fundamental data of the financial market, provides a fine-grained view of market dynamics while poses significant challenges in dealing with the esteemed deep models due to its strong autocorrelation, cross-feature constrains, and feature scale disparity. Existing approaches often tightly couple representation learning with specific downstream tasks in an end-to-end manner, failed to analyze the learned representations individually and explicitly, limiting their reusability and generalization. This paper conducts the first systematic comparative study of LOB representation learning, aiming to identify the effective way of extracting transferable, compact features that capture essential LOB properties. We introduce LOBench, a standardized benchmark with real China A-share market data, offering curated datasets, unified preprocessing, consistent evaluation metrics, and strong baselines. Extensive experiments validate the sufficiency and necessity of LOB representations for various downstream tasks and highlight their advantages over both the traditional task-specific end-to-end models and the advanced representation learning models for general time series. Our work establishes a reproducible framework and provides clear guidelines for future research. Datasets and code will be publicly available at https://github.com/financial-simulation-lab/LOBench.
The field of artificial intelligence (AI) in quantitative investment has seen significant advancements, yet it lacks a standardized benchmark aligned with industry practices. This gap hinders research progress and limits the practical application of academic innovations. We present QuantBench, an industrial-grade benchmark platform designed to address this critical need. QuantBench offers three key strengths: (1) standardization that aligns with quantitative investment industry practices, (2) flexibility to integrate various AI algorithms, and (3) full-pipeline coverage of the entire quantitative investment process. Our empirical studies using QuantBench reveal some critical research directions, including the need for continual learning to address distribution shifts, improved methods for modeling relational financial data, and more robust approaches to mitigate overfitting in low signal-to-noise environments. By providing a common ground for evaluation and fostering collaboration between researchers and practitioners, QuantBench aims to accelerate progress in AI for quantitative investment, similar to the impact of benchmark platforms in computer vision and natural language processing.
Machine learning-based surrogate models have emerged as a powerful tool to accelerate simulation-driven scientific workflows. However, their widespread adoption is hindered by the lack of large-scale, diverse, and standardized datasets tailored to physics-based simulations. While existing initiatives provide valuable contributions, many are limited in scope-focusing on specific physics domains, relying on fragmented tooling, or adhering to overly simplistic datamodels that restrict generalization. To address these limitations, we introduce PLAID (Physics-Learning AI Datamodel), a flexible and extensible framework for representing and sharing datasets of physics simulations. PLAID defines a unified standard for describing simulation data and is accompanied by a library for creating, reading, and manipulating complex datasets across a wide range of physical use cases (gitlab.com/drti/plaid). We release six carefully crafted datasets under the PLAID standard, covering structural mechanics and computational fluid dynamics, and provide baseline benchmarks using representative learning methods. Benchmarking tools are made available on Hugging Face, enabling direct participation by the community and contribution to ongoing evaluation efforts (huggingface.co/PLAIDcompetitions).
The exponential growth of artificial intelligence (AI) applications has exposed the inefficiency of conventional von Neumann architectures, where frequent data transfers between compute units and memory create significant energy and latency bottlenecks. Analog Computing-in-Memory (ACIM) addresses this challenge by performing multiply-accumulate (MAC) operations directly in the memory arrays, substantially reducing data movement. However, designing robust ACIM accelerators requires accurate modeling of device- and circuit-level non-idealities. In this work, we present NeuroSim V1.5, introducing several key advances: (1) seamless integration with TensorRT's post-training quantization flow enabling support for more neural networks including transformers, (2) a flexible noise injection methodology built on pre-characterized statistical models, making it straightforward to incorporate data from SPICE simulations or silicon measurements, (3) expanded device support including emerging non-volatile capacitive memories, and (4) up to 6.5x faster runtime than NeuroSim V1.4 through optimized behavioral simulation. The combination of these capabilities uniquely enables systematic design space exploration across both accuracy and hardware efficiency metrics. Through multiple case studies, we demonstrate optimization of critical design parameters while maintaining network accuracy. By bridging high-fidelity noise modeling with efficient simulation, NeuroSim V1.5 advances the design and validation of next-generation ACIM accelerators. All NeuroSim versions are available open-source at https://github.com/neurosim/NeuroSim.
This paper introduces SecRepoBench, a benchmark to evaluate LLMs on secure code generation in real-world repositories. SecRepoBench has 318 code generation tasks in 27 C/C++ repositories, covering 15 CWEs. We evaluate 19 state-of-the-art LLMs using our benchmark and find that the models struggle with generating correct and secure code. In addition, the performance of LLMs to generate self-contained programs as measured by prior benchmarks do not translate to comparative performance at generating secure and correct code at the repository level in SecRepoBench. We show that the state-of-the-art prompt engineering techniques become less effective when applied to the repository level secure code generation problem. We conduct extensive experiments, including an agentic technique to generate secure code, to demonstrate that our benchmark is currently the most difficult secure coding benchmark, compared to previous state-of-the-art benchmarks. Finally, our comprehensive analysis provides insights into potential directions for enhancing the ability of LLMs to generate correct and secure code in real-world repositories.
Ensuring the generalisability of clinical machine learning (ML) models across diverse healthcare settings remains a significant challenge due to variability in patient demographics, disease prevalence, and institutional practices. Existing model evaluation approaches often rely on real-world datasets, which are limited in availability, embed confounding biases, and lack the flexibility needed for systematic experimentation. Furthermore, while generative models aim for statistical realism, they often lack transparency and explicit control over factors driving distributional shifts. In this work, we propose a novel structured synthetic data framework designed for the controlled benchmarking of model robustness, fairness, and generalisability. Unlike approaches focused solely on mimicking observed data, our framework provides explicit control over the data generating process, including site-specific prevalence variations, hierarchical subgroup effects, and structured feature interactions. This enables targeted investigation into how models respond to specific distributional shifts and potential biases. Through controlled experiments, we demonstrate the framework's ability to isolate the impact of site variations, support fairness-aware audits, and reveal generalisation failures, particularly highlighting how model complexity interacts with site-specific effects. This work contributes a reproducible, interpretable, and configurable tool designed to advance the reliable deployment of ML in clinical settings.
Moral stories are a time-tested vehicle for transmitting values, yet modern NLP lacks a large, structured corpus that couples coherent narratives with explicit ethical lessons. We close this gap with TF1-EN-3M, the first open dataset of three million English-language fables generated exclusively by instruction-tuned models no larger than 8B parameters. Each story follows a six-slot scaffold (character -> trait -> setting -> conflict -> resolution -> moral), produced through a combinatorial prompt engine that guarantees genre fidelity while covering a broad thematic space. A hybrid evaluation pipeline blends (i) a GPT-based critic that scores grammar, creativity, moral clarity, and template adherence with (ii) reference-free diversity and readability metrics. Among ten open-weight candidates, an 8B-parameter Llama-3 variant delivers the best quality-speed trade-off, producing high-scoring fables on a single consumer GPU (<24 GB VRAM) at approximately 13.5 cents per 1,000 fables. We release the dataset, generation code, evaluation scripts, and full metadata under a permissive license, enabling exact reproducibility and cost benchmarking. TF1-EN-3M opens avenues for research in instruction following, narrative intelligence, value alignment, and child-friendly educational AI, demonstrating that large-scale moral storytelling no longer requires proprietary giant models.
Large language models (LLMs) hold great promise for medical applications and are evolving rapidly, with new models being released at an accelerated pace. However, current evaluations of LLMs in clinical contexts remain limited. Most existing benchmarks rely on medical exam-style questions or PubMed-derived text, failing to capture the complexity of real-world electronic health record (EHR) data. Others focus narrowly on specific application scenarios, limiting their generalizability across broader clinical use. To address this gap, we present BRIDGE, a comprehensive multilingual benchmark comprising 87 tasks sourced from real-world clinical data sources across nine languages. We systematically evaluated 52 state-of-the-art LLMs (including DeepSeek-R1, GPT-4o, Gemini, and Llama 4) under various inference strategies. With a total of 13,572 experiments, our results reveal substantial performance variation across model sizes, languages, natural language processing tasks, and clinical specialties. Notably, we demonstrate that open-source LLMs can achieve performance comparable to proprietary models, while medically fine-tuned LLMs based on older architectures often underperform versus updated general-purpose models. The BRIDGE and its corresponding leaderboard serve as a foundational resource and a unique reference for the development and evaluation of new LLMs in real-world clinical text understanding. The BRIDGE leaderboard: https://huggingface.co/spaces/YLab-Open/BRIDGE-Medical-Leaderboard
Autonomous UI agents powered by AI have tremendous potential to boost human productivity by automating routine tasks such as filing taxes and paying bills. However, a major challenge in unlocking their full potential is security, which is exacerbated by the agent's ability to take action on their user's behalf. Existing tests for prompt injections in web agents either over-simplify the threat by testing unrealistic scenarios or giving the attacker too much power, or look at single-step isolated tasks. To more accurately measure progress for secure web agents, we introduce WASP -- a new publicly available benchmark for end-to-end evaluation of Web Agent Security against Prompt injection attacks. Evaluating with WASP shows that even top-tier AI models, including those with advanced reasoning capabilities, can be deceived by simple, low-effort human-written injections in very realistic scenarios. Our end-to-end evaluation reveals a previously unobserved insight: while attacks partially succeed in up to 86% of the case, even state-of-the-art agents often struggle to fully complete the attacker goals -- highlighting the current state of security by incompetence.
Large Language Models (LLMs) have emerged as personalized assistants for users across a wide range of tasks -- from offering writing support to delivering tailored recommendations or consultations. Over time, the interaction history between a user and an LLM can provide extensive information about an individual's traits and preferences. However, open questions remain on how well LLMs today can effectively leverage such history to (1) internalize the user's inherent traits and preferences, (2) track how the user profiling and preferences evolve over time, and (3) generate personalized responses accordingly in new scenarios. In this work, we introduce the PERSONAMEM benchmark. PERSONAMEM features curated user profiles with over 180 simulated user-LLM interaction histories, each containing up to 60 sessions of multi-turn conversations across 15 real-world tasks that require personalization. Given an in-situ user query, i.e. query issued by the user from the first-person perspective, we evaluate LLM chatbots' ability to identify the most suitable response according to the current state of the user's profile. We observe that current LLMs still struggle to recognize the dynamic evolution in users' profiles over time through direct prompting approaches. As a consequence, LLMs often fail to deliver responses that align with users' current situations and preferences, with frontier models such as GPT-4.1, o4-mini, GPT-4.5, o1, or Gemini-2.0 achieving only around 50% overall accuracy, suggesting room for improvement. We hope that PERSONAMEM, along with the user profile and conversation simulation pipeline, can facilitate future research in the development of truly user-aware chatbots. Code and data are available at github.com/bowen-upenn/PersonaMem.
Fish Feeding Intensity Assessment (FFIA) is crucial in industrial aquaculture management. Recent multi-modal approaches have shown promise in improving FFIA robustness and efficiency. However, these methods face significant challenges when adapting to new fish species or environments due to catastrophic forgetting and the lack of suitable datasets. To address these limitations, we first introduce AV-CIL-FFIA, a new dataset comprising 81,932 labelled audio-visual clips capturing feeding intensities across six different fish species in real aquaculture environments. Then, we pioneer audio-visual class incremental learning (CIL) for FFIA and demonstrate through benchmarking on AV-CIL-FFIA that it significantly outperforms single-modality methods. Existing CIL methods rely heavily on historical data. Exemplar-based approaches store raw samples, creating storage challenges, while exemplar-free methods avoid data storage but struggle to distinguish subtle feeding intensity variations across different fish species. To overcome these limitations, we introduce HAIL-FFIA, a novel audio-visual class-incremental learning framework that bridges this gap with a prototype-based approach that achieves exemplar-free efficiency while preserving essential knowledge through compact feature representations. Specifically, HAIL-FFIA employs hierarchical representation learning with a dual-path knowledge preservation mechanism that separates general intensity knowledge from fish-specific characteristics. Additionally, it features a dynamic modality balancing system that adaptively adjusts the importance of audio versus visual information based on feeding behaviour stages. Experimental results show that HAIL-FFIA is superior to SOTA methods on AV-CIL-FFIA, achieving higher accuracy with lower storage needs while effectively mitigating catastrophic forgetting in incremental fish species learning.
Positional embeddings (PE) play a crucial role in Vision Transformers (ViTs) by providing spatial information otherwise lost due to the permutation invariant nature of self attention. While absolute positional embeddings (APE) have shown theoretical advantages over relative positional embeddings (RPE), particularly due to the ability of sinusoidal functions to preserve spatial inductive biases like monotonicity and shift invariance, a fundamental challenge arises when mapping a 2D grid to a 1D sequence. Existing methods have mostly overlooked or never explored the impact of patch ordering in positional embeddings. To address this, we propose LOOPE, a learnable patch-ordering method that optimizes spatial representation for a given set of frequencies, providing a principled approach to patch order optimization. Empirical results show that our PE significantly improves classification accuracy across various ViT architectures. To rigorously evaluate the effectiveness of positional embeddings, we introduce the "Three Cell Experiment", a novel benchmarking framework that assesses the ability of PEs to retain relative and absolute positional information across different ViT architectures. Unlike standard evaluations, which typically report a performance gap of 4 to 6% between models with and without PE, our method reveals a striking 30 to 35% difference, offering a more sensitive diagnostic tool to measure the efficacy of PEs. Our experimental analysis confirms that the proposed LOOPE demonstrates enhanced effectiveness in retaining both relative and absolute positional information.
Restoring any degraded image efficiently via just one model has become increasingly significant and impactful, especially with the proliferation of mobile devices. Traditional solutions typically involve training dedicated models per degradation, resulting in inefficiency and redundancy. More recent approaches either introduce additional modules to learn visual prompts, significantly increasing model size, or incorporate cross-modal transfer from large language models trained on vast datasets, adding complexity to the system architecture. In contrast, our approach, termed AnyIR, takes a unified path that leverages inherent similarity across various degradations to enable both efficient and comprehensive restoration through a joint embedding mechanism, without scaling up the model or relying on large language models.Specifically, we examine the sub-latent space of each input, identifying key components and reweighting them first in a gated manner. To fuse the intrinsic degradation awareness and the contextualized attention, a spatial-frequency parallel fusion strategy is proposed for enhancing spatial-aware local-global interactions and enriching the restoration details from the frequency perspective. Extensive benchmarking in the all-in-one restoration setting confirms AnyIR's SOTA performance, reducing model complexity by around 82\% in parameters and 85\% in FLOPs. Our code will be available at our Project page (https://amazingren.github.io/AnyIR/)
High-fidelity simulation is essential for robotics research, enabling safe and efficient testing of perception, control, and navigation algorithms. However, achieving both photorealistic rendering and accurate physics modeling remains a challenge. This paper presents a novel simulation framework--the Unreal Robotics Lab (URL) that integrates the Unreal Engine's advanced rendering capabilities with MuJoCo's high-precision physics simulation. Our approach enables realistic robotic perception while maintaining accurate physical interactions, facilitating benchmarking and dataset generation for vision-based robotics applications. The system supports complex environmental effects, such as smoke, fire, and water dynamics, which are critical for evaluating robotic performance under adverse conditions. We benchmark visual navigation and SLAM methods within our framework, demonstrating its utility for testing real-world robustness in controlled yet diverse scenarios. By bridging the gap between physics accuracy and photorealistic rendering, our framework provides a powerful tool for advancing robotics research and sim-to-real transfer.
Recent advancements in Large Vision-Language Models (LVLMs) have demonstrated remarkable multimodal perception capabilities, garnering significant attention. While numerous evaluation studies have emerged, assessing LVLMs both holistically and on specialized tasks, fine-grained image tasks-fundamental to computer vision-remain largely unexplored. To fill this gap, we introduce a comprehensive fine-grained evaluation benchmark, i.e., FG-BMK, comprising 1.01 million questions and 0.33 million images. Our evaluation systematically examines LVLMs from both human-oriented and machine-oriented perspectives, focusing on their semantic recognition and fine-grained feature representation capabilities. Through extensive experiments on twelve representative LVLMs/VLMs, we uncover key findings regarding the influence of training paradigms, modality alignment, perturbation susceptibility, and fine-grained category reasoning on task performance. This work provides critical insights into the limitations of current LVLMs and offers guidance for future data construction and model design in the development of more advanced LVLMs. Our code is open-source and available at https://github.com/SEU-VIPGroup/FG-BMK.