New
Everything new in AI as it appears: launches announced by the labs, brand-new models, datasets, papers and open-source tools. Checked every minute or two. Launches feed · Trending · News
Live now
- Connecting…
Saturday 24 October
- DatasetMalavnaik/finfacts
FinFacts: SEC fundamentals eval set and 10-K chunks Built by FinFacts, a pipeline that turns SEC EDGAR XBRL filings for the S&P 500 into certified, point-in-time-correct
- Datasetkiroaiseoul/replay_base_chain_1025b
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "robot_type": "mobileai_robot", "total_episodes": 1, "total_frames
- Datasetdarwinagent/WIA-Corpus-SFT
WIA-Corpus-SFT 文件用途与样本数 每行是一条样本，每条样本包含一组有序消息。 文件 用途 样本数 顶层字段数 train.jsonl 世界模型监督微调训练集 9,181 3 字段说明 字段 类型 用途 messages 数组 按顺序保存样本中的对话消息。 domain 字符串 任务领域，如 swe、terminal、web。
- Dataset
- Dataset
- Datasetarmnet/eval_busybox_push_green_button_diffusion_18h34m_24-oct-2026
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "robot_type": "so-101", "total_episodes": 1, "total_frames": 279,
- Datasethirundo-io/spml-prompt-injection-refusals-strict-answered-refusal-only
spml-prompt-injection-refusals-strict-answered-refusal-only hirundo-io/spml-prompt-injection-refusals-strict-answered at revision 6561ab2329b673bf9c37bc85da91abd61c1df44b
- Datasetxmarva/tfs-translations-en-ru
Translations of the course's English–Russian test set, with COMET scores Made in lesson 4 of the course "Transformers: from scratch to production": https://github.com/xma
- DatasetRafaelJaime/smartseed-grasp-b-merged
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 30, "features": { "action": { "dtype": "float32", "names":
- DatasetBeaconf88/TACO
TACO is a benchmark for Python code generation, it includes 25443 problems and 1000 problems for train and test splits.
- Dataset
- DatasetLunita02/Cubos_200
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 30, "features": { "action": { "dtype": "float32", "names":
- DatasetRafaelJaime/smartseed-grasp-12s-clean
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 30, "features": { "action": { "dtype": "float32", "names":
- Dataset
- DatasetLunita02/Cubos_50_3
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 30, "features": { "action": { "dtype": "float32", "names":
- Datasetxdotli/oea-safecase-20261024-dfd597
SafeCase: six local C regression-test tasks Six new, deterministic, self-contained toy C fixture pairs. The agent submits a bounded decimal integer as a test input. Rewar
- DatasetAarjan/swar-neptts-bench
Swar on NepTTS-Bench Audio from Swar (Naamche Labs) for the 205 scored NepTTS-Bench sentences, named sent_XXX.wav as the benchmark expects. One take per sentence, no sele
- Dataset
- DatasetNelvoraAI/russian-math-dialogues
🧮 Russian Math Dialogues 11 277 синтетических диалогов по математике — полностью на русском языке ∑ Считаем, доказываем, объясняем — по-русски ∫ 📖 О датасете Russian Ma
- Datasetschraadsweather/nws-text
NWS text products Deduplicated NWWS-OI/EMWIN text products, verbatim, with parsed metadata (PIL, office, VTEC events, UGC zones, category, severity tier, expiry). Layout:
- DatasetPureOne/EIQ-Lambda
EIQ-Λ v2 The Fractal Holography of Physical Objects - GPU-Accelerated Inverse Scattering, Programmable Matter, and Recursive Structural Intelligence. AI Assisted Universa
- Datasetjalenluorion/prefetchable-store-corpus-v4
Pre-tokenised language-model training corpus, v4 The v4 build of the v3 corpus: a byte-identical superset of v3 for longer runs. 64,600,000,000 tokens of FineWeb-Edu, tok
- Datasetva1bhavagrawa1/data_mix_v4_3d_wireframes
Data Mix v4 This is a 60,000-record paired image-editing dataset packaged as 1,200 tar shards (50 records per shard). manifest.jsonl indexes each record and its six membe
- DatasetNelvoraAI/GPT-6-Luna-RU
GPT-6 Luna General Purpose (RU) Синтетический русскоязычный диалоговый датасет общего назначения (General Purpose) для supervised fine-tuning языковых моделей. Все 6 015
- DatasetNelvoraAI/RuThink-OSS-120B
🧠 RuThink-OSS-120B Русский SFT-датасет с цепочками рассуждений: истории + медицина Диалоговый датасет для дообучения языковых моделей на русском языке. Все 3 220 примеро
- Dataset
- Dataset
- DatasetAgenticFilmLab/minimax-h3-ptd-spark-comparison-record
MiniMax H3: PTD4, PTD8 and Spark comparison record Public evidence for the Agentic Film Lab episode approved as v015, 8:29.625. This is a small documented experiment, not
- Dataset
- DatasetLunita02/hackathon_dataset_20261024_200148
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 30, "features": { "action": { "dtype": "float32", "names":
- Dataset
- Dataset
- Dataset
- Datasetar-mine/e5-c3-code
E5+C3 original training code and evaluation documentation Public release, 2026-10-25 (Asia/Singapore). This Hugging Face dataset hosts a Git source repository, cloneable
- Dataset
- DatasetLunita02/hackathon_dataset_20261024_195012
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 30, "features": { "action": { "dtype": "float32", "names":
- Dataset
- Dataset
- DatasetRevanthGundala/026-pick-place-soda-can-50-trimmed
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 30, "features": { "action": { "dtype": "float32", "shape":
- DatasetPokepika/dolma
Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Research
Headlines and links come from the labs' own blogs, Hacker News and the GDELT news index, and from the sources our index already follows. We show a short excerpt and always link to the original. A launch reported by a news site rather than the lab is marked “Reported” until the lab itself announces it.