1
我国首版全国几何基准影像成果发布，定位精度提升 1 倍以上
IT之家 8 月 28 日消息，据自然资源部官方公众号，在 8 月 27 日自然资源部召开的例行新闻发布会上，我国首版全国几何基准影像成果正式发布。该项成果是自然资源部积极服务数字中国建设、推动测绘地理信息事业转型升级的重大标志性成果，也是面向社会服务的测绘地理信息新型公共产品。 IT之家从官方介绍…
2
谷歌推出全球首个双盲 AI 评估技术，基于加密环境保障基准测试公正性
IT之家 8 月 27 日消息，谷歌宣布推出 全球首个针对前沿专有 AI 模型的双盲评估 ，将外部评估限制在加密“黑箱”环境中，避免模型提前获取测试信息来优化性能。 就像学生考前不能提前看到试题才能真实反映水平一样，当前 AI 模型评估也面临同样问题：如果模型提前接触测试题目（IT之家注：也就是所谓…
3
OpenAI’s Jalapeño chip is built for fast inference at scale, benchmarks show
OpenAI自研Jalapeño芯片基准测试曝光，专为大规模快速推理而生，直面英伟达Blackwell。
Tested on SemiAnalysis’ InferenceX benchmark, Jalapeño registered both more tokens per user and more throughput per kilowatt than the currently availa…
4
ATP-Bench: Towards Agentic Tool Planning for MLLM Interleaved Generation
多模态大模型如何自主规划工具生成图文交织内容？首个基准ATP-Bench来了。
arXiv:2603.29902v2 Announce Type: replace Abstract: Interleaved text-and-image generation represents a significant frontier for Multimodal Large Langu…
5
MedReaMM: Evaluating Large Multimodal Models on Expert-Level Clinical Diagnostic Synthesis
破解AI看病难题：全新基准MedReaMM如何考核多模态大模型的专业诊断整合能力
arXiv:2608.22323v1 Announce Type: new Abstract: The application of Large Language Models (LLMs) to diagnostic decision-making has garnered growing int…
6
Omni-SafetyBench: A Benchmark for Safety Evaluation of Audio-Visual Large Language Models
首个聚焦音视频大模型安全性的基准，曝光跨模态越狱漏洞，为多模态安全研究提供关键测试标准。
arXiv:2508.07173v3 Announce Type: replace Abstract: Omni-modal Large Language Models (OLLMs) that integrate visual, auditory, and textual processing f…
7
Show HN: Poka-Yoke – mistake-proofing skills for Claude Code, with a benchmark
为Claude Code打造防错技能，经591次盲测验证，六类模型均显著提升。
Article URL: https://github.com/rainmanjam/poka-yoke Comments URL: https://news.ycombinator.com/item?id=49427559 Points: 1 # Comments: 0
8
Beyond RGB: Benchmarking and Enhancing MLLMs for Hyperspectral Image Understanding via Training-Free Reasoning Framework
多模态大模型不止看懂RGB，高光谱成像理解迎来全新基准与免训练推理框架。
arXiv:2604.08884v2 Announce Type: replace-cross Abstract: Multimodal Large Language Models (MLLMs) have achieved strong performance on RGB image under…
9
Show HN: CookWing, an AI chef that doesn't hallucinate quantities
AI厨师不瞎编调料量？CookWing用溯源锚定表格解决份量幻觉，还自建基准测试。
Hey HN! I just released Cookwing. This is an AI app that solves the biggest problem I had when I was cooking with ChatGPT : The quantities are often h…
10
Stress Testing Unlearning Algorithms
系统检验机器遗忘算法在极端压力下的鲁棒性，为隐私保护研究划出新基准。
arXiv:2608.22527v1 Announce Type: new Abstract: Recently, machine unlearning, the removal of specific training data influence from a model, has gained…
11
Benchmarking Patent Drafting from Inventor-Style Disclosures
科研新基准：从发明人式披露生成专利文本，直击大模型专业写作能力短板。
arXiv:2608.21249v1 Announce Type: new Abstract: While recent large language models (LLMs) have achieved promising results on individual patent draftin…
12
I Shouldn't Need an LLM to Explain My LLM
揭秘AI圈“自我舔舐”陷阱：用LLM解释LLM，代码生成的测试金标准还能可信吗？
Article URL: https://daviesgeek.com/I-Shouldn%E2%80%99t-Need-an-LLM-to-Explain-My-LLM Comments URL: https://news.ycombinator.com/item?id=49409282 Poin…
13
Nvidia just showed that the harness, not the AI model, is now the real hero
别只盯着模型参数，一个调教得当的“缰绳”竟能让Claude Opus 5在推理基准上满分通关。
Nvidia research shows that AI agents can perform well, and not go off the deep end, through fine-tuning, even if the AI model isn't that great at the …
14
I benched frontier AI LLM models on politics, ethics, and personality traits
前沿AI模型的政治立场、伦理取向与人格特质，一次基准测试看透底牌
Article URL: https://www.blackbench.ai/ Comments URL: https://news.ycombinator.com/item?id=49382540 Points: 3 # Comments: 2
15
PersonalBench: Measuring the Authorship Gap in LLM Personalization
不看任务准确率，专测AI模仿你文风的“神似度”，PersonalBench填补了个性化写作评测空白
arXiv:2608.19746v1 Announce Type: new Abstract: Personalized text generation aims to make LLMs write in a specific individual's style, yet existing be…
16
AI4AI-Bench: Benchmarking LLM Agents in Algorithmic Design for Recursive Self-Improvement
首个面向递归自我改进的AI智能体算法设计基准，直击AI自我进化核心命题。
arXiv:2608.20318v1 Announce Type: cross Abstract: Recursive self-improvement (RSI) asks whether an AI system can improve the process that produces AI …
17
MemTrapBench: Benchmarking Cognitive Traps in LLM Memory Use
大模型记忆中的认知陷阱有了专门的评测基准，帮你识别AI“想当然”的隐患。
arXiv:2608.20202v1 Announce Type: cross Abstract: Memory has become a key component of large language models, enabling them to retain information and …
18
阿里发布 Qwen-UI-Agent，主打让模型真正“会用”每一块屏幕
阿里Qwen-UI-Agent让AI真正操作手机屏幕，真机基准超越多款国际大模型
IT之家 8 月 20 日消息，据千问大模型官方公众号，阿里巴巴正式推出 Qwen-UI-Agent ，一个以真实世界为中心的 GUI 智能体基座模型，涵盖移动端、电脑端、网页端以及深度搜索（DeepSearch）环境。 据介绍，Qwen-UI-Agent 在 GUI 任务上全面对标乃至超越业界旗舰…
19
高通调整骁龙 C 处理器幻灯片，移除 2 项测试结果
高通紧急修改骁龙C处理器宣传幻灯片，悄然删除两项基准测试结果，背后有何考量？
IT之家 8 月 20 日消息，科技媒体 Tom's Hardware 昨日（8 月 19 日）发布博文，报道称高通已联系多家媒体，要求改用新版骁龙 C 处理器（Snapdragon C）幻灯片， 主要删除了“空闲应用”和“网页浏览”2 项基准测试结果。 IT之家曾于 8 月 13 日报道， 高通披…
20
Fair ASR: Re-Evaluating Black-Box Jailbreaks under Shared Target-Call Budgets
挑战黑盒越狱评估的公平性，引入共享调用预算新框架，重新定义攻击成功率，值得AI安全研究者细读。
arXiv:2608.17360v1 Announce Type: cross Abstract: Reliable jailbreak evaluation is essential for assessing LLM safety, but most existing studies rely …