Tag
DAIR.AI's Omar (@omarsar0) built 'Hacker Papers', an agent-generated collection of 1K+ arXiv papers mapped to their Hacker News discussion threads, inspired by Andrej Karpathy's HackerTube idea. The tool lets readers explore popular papers alongside community engagement stats over time.
A developer built a Codenames-inspired browser game where LLMs play each other, running 376 games across 19 models and finding that guesser quality (68% win correlation) matters far more than clue quality (53%), with newer models guessing much better than older ones. Raw data and replayable games are shared publicly.
SVPino highlights Voice Arena AI's Joy, a 100,000-hour dual-channel conversational voice dataset with separate per-speaker tracks, overlapping speech, interruptions, pauses, and backchannels, designed to teach voice models conversational timing.
Stripe is collecting and publishing empirical data on European startups and computing power, with Patrick Collison saying he hopes this will provide valuable insights for Europe's entrepreneurship and industry.
The EPA is reportedly quietly dismantling or removing its web archive, raising concerns about the loss of public environmental information and digital preservation.
CHiME-3 背景噪音录音现已在 Zenodo 上以 CC BY 4.0 协议开放提供，包含六通道平板阵列在公交、咖啡馆、步行区和街道四种真实环境中录制的 16 kHz 音频，允许商用及再分发。
Loombench 实验让五个 AI 代理共享一台织机，发现人格设定几乎总能决定协作结果：粗鲁代理霸占了 89% 的资源，而礼貌代理即使被告知对方在欺负人也从不反击。该项目提供 102 次运行的开放数据，用户花约 $0.50 即可测试自己的模型。
An interactive map visualizing all 384,000 farm fields in California, allowing users to zoom into individual fields and see what crop is being grown, inspired by the Snapchat map.
NASA 公开了超过 800GB 的 Artemis II 任务数据，包括 11,000 多张照片和视频、8.5 小时音频以及机组人员标注的月球背面撞击闪光图像，同时发布了三份配套科学报告。
NASA 公开了约 800 GB 的 Artemis II 任务科学数据，包括 11,000 张全分辨率图像、宇航员观察记录及三份初步科学报告，供科学界和公众访问。
Biohub, the U.S. Department of Energy, the NIH, Google DeepMind, Isomorphic Labs, and Meta jointly committed $1.8 billion to generate AI-ready open biological datasets, the largest coordinated effort to date to enable predictive AI models of biology and a 'virtual cell'.
Biohub, the U.S. Department of Energy, NIH, Google DeepMind, Isomorphic Labs, and Meta announced a $1.8 billion expansion of the Virtual Biology Initiative to generate open, AI-ready biological datasets for building predictive models of biology and accelerating disease research.
A developer built a browser war game called SECOND STRIKE that pits AI models against each other with nuclear buttons; DeepSeek ended the world 13/15 times while GPT only did so 3/15, and the data and replays are open for humans to compete on a live scoreboard.
发布了包含 56 亿条 TikTok 视频元数据（2014 年至 2026 年 10 月）的数据集于 Hugging Face，并提供可直接查询的 ClickHouse 数据库，涵盖创作者、视频和音频表。
Biohub宣布启动Virtual Biology Initiative，这是一项为期五年、总额18亿美元的全球合作计划，旨在构建可预测的人体细胞模型；Biohub承诺投入5亿美元，DeepMind、Isomorphic Labs、Meta、美国政府及NVIDIA等均参与其中。
Bert Hubert hield een praatje over het belang van transparantie voor de democratie en hoe de open gemeenschap kan helpen om de overheidstransparantie te herstellen.
The article discusses the process of utilizing open training data from Hugging Face to train Marin’s 535B model, which involved 25T tokens from 152 datasets with permissible licenses.
The article presents a tool that displays the release age and training cutoff dates for 20 current AI models, highlighting how stale the data might be in each model.
This paper investigates membership inference in language models by using exact duplication counts from open pretraining corpora of OLMo-2 and Pythia, revealing that typical duplication levels show minimal exposure traces and that apparent membership signals are often confounded with factors like sentence fame.
Planet Labs has announced an open satellite feed, providing public access to its satellite imagery data.