独立 AI 资讯库
从自动抓取数据库实时读取，适合快速浏览、检索和后续扩展成完整资讯站。
arXiv:2608.14577v1 Announce Type: new Abstract: Frontier large language models (LLMs) safety evaluation has largely treated harmful generation as an attack outcome rather than as an object of analysis. Consequently, little is known about the harmful outputs produced during model misbehavior, partly
arXiv:2608.14551v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used for title-abstract screening in systematic reviews, but their decisions lack calibrated uncertainty. We show that an auxiliary BERT+GCN classifier supplies a structured uncertainty signal that improve
arXiv:2608.14667v1 Announce Type: new Abstract: Large language model-based agents are increasingly deployed as collaborators in scientific discovery yet most current work focuses on the autonomous capabilities of "AI Scientists". We argue that this overlooks the social aspects of scientific teamwor
arXiv:2608.14651v1 Announce Type: new Abstract: Effective disaster risk communication is a foundational humanitarian challenge, yet current emergency infrastructure fails to meet the needs of individuals with access and functional needs, including hard-of-hearing individuals, pregnant women, mother
arXiv:2608.14641v1 Announce Type: new Abstract: Agentic systems increasingly delegate model selection to a router, yet open-source routers are usually evaluated with different tasks, candidate pools, and execution protocols, limiting direct comparison. We present a common measurement protocol and h
arXiv:2608.14624v1 Announce Type: new Abstract: Multi-agent LLM systems have emerged as an important deployment paradigm for AI services, where each user request is decomposed into a sequence of specialized agents. Across these workflows, every agent repeatedly executes a fixed context consisting o
arXiv:2608.14615v1 Announce Type: new Abstract: Large Language Models (LLMs) perform well on established code-generation and mathematical-reasoning benchmarks, but their capabilities in mechanics and spatial geometry, here denoted as mechanical engineering awareness, has not been quantified systema
arXiv:2608.14613v1 Announce Type: new Abstract: Modern LLM-agent frameworks increasingly interoperate through standards such as Anthropic's Model Context Protocol (MCP) for agent-to-tool access and Google's Agent2Agent (A2A) protocol for agent delegation and negotiation. However, these protocols sp
arXiv:2608.14610v1 Announce Type: new Abstract: Legal reasoning tasks such as legal judgment prediction (LJP) require identifying the temporally correct version of the law governing a case -- a capability we term temporal applicable-law determination. However, whether large language models (LLMs) c
arXiv:2608.14590v1 Announce Type: new Abstract: LLM agents increasingly perform irreversible real-world actions, including database updates, API calls, file operations, and autonomous use of tools. However, no existing system provides formally grounded, task-level safety guarantees for the plans th
arXiv:2608.14588v1 Announce Type: new Abstract: Sequential multi-agent LLM pipelines chain specialized agents without verification at handoffs, creating a structural flaw with measurable and severe consequences. We show that hallucinations injected at Stage 1 do not merely persist; they transform:
arXiv:2608.14587v1 Announce Type: new Abstract: Background: Recent advances in information retrieval (IR) leverage both dense and sparse representations, large language models (LLMs), and specialized retrieval models to improve ranking accuracy, relevance, and cross-lingual performance. Complementa
arXiv:2608.14585v1 Announce Type: new Abstract: Euclidean geometry is a compelling testbed for AI reasoning, as it demands the combination of intuitive diagram understanding, axiomatic deduction, and algebraic computation. Yet, existing approaches typically address only a subset of these abilities
arXiv:2608.14580v1 Announce Type: new Abstract: OGX (Open GenAI Stack) is an open-source AI application server and Python library that implements the APIs of major frontier labs (OpenAI, Anthropic, Google) with pluggable backend providers. Developers building agentic AI applications--such as retrie
arXiv:2608.14579v1 Announce Type: new Abstract: Logic synthesis optimization poses significant challenges due to exponentially growing search spaces, sparse reward signals, and diverse logic structures. Traditional expert-designed flows lack adaptability, while reinforcement learning (RL) methods o
arXiv:2608.14569v1 Announce Type: new Abstract: Neural solvers for constraint satisfaction problems have achieved remarkable in-distribution accuracy, yet they suffer from a fundamental limitation persistent constraint violations occur under distribution shifts even when the model reports high conf
arXiv:2608.14566v1 Announce Type: new Abstract: Recent work on evaluating the moral competence of large language models (LLMs) has focused primarily on what we call the moral value problem, i.e., whether model outputs align with human moral values. In contrast, the moral norm problem, i.e., whether
arXiv:2608.14559v1 Announce Type: new Abstract: Effective communication in multi-agent reinforcement learning requires agents to decide not only \textit{what} to communicate, but when? Existing approaches either communicate at every timestep or learn a binary gate through REINFORCE policy gradients
arXiv:2608.14558v1 Announce Type: new Abstract: Current multimodal models have demonstrated remarkable proficiency in recognizing static visual and auditory content. However, their capacity for abstract perceptual reasoning, inferring unseen information from dynamic, generative processes, remains a
arXiv:2608.14552v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly evaluated and used in medicine, but clinical usefulness depends on answer accuracy and whether confidence tracks evidence quality and uncertainty. We developed a controlled, psychophysics-inspired clinical
处方转录错误在临床影响上无法等同视之，模型虚构药品与误读清晰剂量的临床风险截然不同，但现有研究通常仅采用单一混合指标评估准确率。RxScribe Bench 针对印度门诊处方场景构建，是一个用于全面评估视觉语言模型（VLM）表现的多维度基准测试。该成果为开发者和医疗 AI 领域提供了细粒度的临床安全与识别能力评估工具，有助于推动医疗文档智能处理的实际落地。
针对时尚属性提取领域评估标准不一致的痛点，该研究提出了四轨评估基准 MODA 通用属性套件（MODA General Attribute Suite）。当前业界评测常将不同难度的图像类型简单聚合成单一分数，并将图像中不可见的字段直接按常规负样本计算，忽视了词表设计带来的偏差。MODA 通过细分四个评估赛道，为时尚领域的多模态属性提取与电商大模型提供了标准化、高精度的评测基准，有助于开发者更准确地诊断模型缺陷与真实泛化能力。
研究团队针对月球科学领域发布了基准数据集 SomBench，旨在加速机器学习在该领域的应用与发展。该数据集整合了月球勘测轨道飞行器（LRO）、月亮女神号（Kaguya/SELENE）、重力恢复与内部结构实验室（GRAIL）及月球探勘者号等探测任务的多载荷观测数据。针对多源任务在空间采样和投影格式上的异质性瓶颈，该基准为研究人员提供了标准化的评估环境，对利用 AI 推进月球地质与空间科学研究具有重要意义。
该论文聚焦年度报告、尽调材料和信息图仪表盘等长文档中的视觉化数字信息检索与问答。作者指出，坐标轴、表格网格和脚注等关键信息常嵌入页面图像中，传统 OCR 流水线会将其扁平化处理，而页面级视觉检索器也往往无法区分不同的上下文示例。论文提出“能力路由视觉检索”和“证据串联”方法，用于改进长上下文文档问答中的证据定位与推理。该研究对需要从复杂企业文档中准确提取数字、图表和注释信息的开发者与行业应用具有参考意义。
针对大型视觉语言模型常将科学图表当作自然照片处理、易产生未对齐数值幻觉的问题，研究团队提出了 STEER 框架。该架构冻结了 Llama-3.2-Vision 编码器，并引入图表结构图编码器（CSGE）、证据锚定单步推理（EASR）以及弱解析器-强推理器对齐（WPSR）三个模块，强制推理链的每一步算术操作都必须引用图表结构节点。评测显示，STEER 在 ChartQA 上取得了 82.70 的平均松弛准确率，显著超越 ChartGemma，并在杜绝 OCR 捷径的 CharXiv 和 ChartQAPro 等复杂基准中展现出更明显的领先优势。该工作为开发者和学术界在多模态 Agent 进行高精度科学文档解析与数据推理方面提供了可信度更高的解决方案。
BEACON 是一种轻量级的特定人物视频生成框架，旨在同时保留人物的视觉身份和个体化面部表情行为。该方法使用参考图像编码身份，并使用参考视频捕捉人物特有的面部动态，从而将外观与表达行为解耦。实验显示，在 MEAD 和 RAVDESS 数据集上，BEACON 仅需微调约 2,000 个样本对、更新预训练 Wan 视频扩散模型约 1% 的参数，即可较现有先进方法提升面部表情表现力，同时维持具有竞争力的身份保持能力。该方法还支持身份与表情的迁移，为更可控、更具个性化的视频生成提供了新的实现路径。
本文提出了一种基于离散 EventGraph、连续 EventField 以及人类可读 EventGlyph 视图构建的结构化时序视频推理管线。在 EPIC-KITCHENS 校准子集的 10 个视频及 50 个时序推理问题评测中，EventField+Glyph 方案取得了 0.98 的总体准确率，相比字幕基线提升 0.40，相较直接使用 VLM 问答提升 0.20。该研究展示了结构化时序表征在保留符号结构与捕捉时间连续性方面的显著优势，不仅提升了视频推理性能，还为多模态长视频理解与诊断提供了兼具性能与可解释性的新范式。
该论文对多模态面部状态分析的任务、方法与资源进行综合梳理。面部状态分析对于理解人类表情、心理建模和人机交互具有重要作用。论文指出，传统基于单一视觉模态的方法往往容易受到环境因素影响，且可解释性较弱；这表明融合多种模态的信息可能对提升相关系统的鲁棒性与可解释性具有意义。
该论文研究多模态大语言模型（MLLM）如何理解鸭兔图等双稳态图像。此类图像可支持多个彼此不兼容的解释，而人类通常一次只会报告其中一种感知结果。作者据此考察 MLLM 是否表现出与人类相似的图像解释与报告特征。该研究可能有助于评估多模态模型的视觉感知、歧义处理与类人认知能力。
该论文关注多模态大语言模型（MLLM）在长视频理解中的关键帧选择问题。由于视觉 token 预算和计算资源有限，MLLM 无法处理长视频中的每一帧。论文评估了与具体 MLLM 无关、可即插即用的关键帧选择方法，旨在为长视频任务中的高效视频采样与模型部署提供参考。原文提供的摘要片段未包含具体实验数据、评测结论或方法细节。