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…
Friday 23 October
- DatasetMikelRubi/so101-pick-cube_20261023_213054
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 10, "features": { "action": { "dtype": "float32", "names":
- Datasetmachinekind/polish-synthetic-speech-voxcpm2-v2
Polish synthetic speech v2 — conversational, VoxCPM2, 2 voices 19646 clips, 22.8 h of modern spoken Polish as a humanoid robot says it: reactions, questions, confirmation
- DatasetIcey444/img2img-short
img2img short-answer cards 52,278 paired 1024×1024 input card → output card examples whose answer is only 1-3 characters: a multiple-choice letter, a yes/no, or a number.
- DatasetSOTAagi2030/Museum-Light-Exposure-Excursions
Museum Light Exposure Excursions This de-identified conservation release covers 2026-09-14. Release totals Public galleries: 4 Accepted readings: 9 Excursion readings: 7
- DatasetMikelRubi/up-down_20261023_212336
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 10, "features": { "action": { "dtype": "float32", "names":
- DatasetMikelRubi/so101-pick-cube_20261023_212258
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 10, "features": { "action": { "dtype": "float32", "names":
- DatasetAeliaTR/aelia-stage8-veri
AELIA Stage 8 - AeLLM 1.08 Ajan verisi AeliaTR ajan modeli (AeLLM 1.08) icin iki dosya: dosya satir rol STAGE_8_AGENT_MASTER.jsonl 15000 SFT QLoRA STAGE_8_GRPO_VERI.jsonl
- DatasetMikelRubi/up-down_20261023_211821
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 10, "features": { "action": { "dtype": "float32", "names":
- DatasetMikelRubi/so101-pick-cube_20261023_211739
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 10, "features": { "action": { "dtype": "float32", "names":
- Dataset
- Datasetsondalex/university-iowa-guitar-samples
University of Iowa Guitar Samples Individual guitar note recordings derived from the University of Iowa Electronic Music Studios' Musical Instrument Samples (MIS) guitar
- Dataset
- Dataset
- Dataset
- Dataset
- Dataset
- Dataset
- DatasetSOTAagi2030/Harbor-Buoy-Alert-Episodes-271
Harbor buoy alert release: harbor-watch-2026-04-18 Observation cutoff: 2026-04-18T12:00:00Z Measure Count Catalog stations 5 Active stations 4 Active stations in included
- DatasetMikelRubi/up-down_20261023_211221
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 10, "features": { "action": { "dtype": "float32", "names":
- DatasetMikelRubi/so101-pick-cube_20261023_211050
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 10, "features": { "action": { "dtype": "float32", "names":
- DatasetMJPansa/qwen38-eagle4-native-features-37m-20261024-5277c167
Qwen3.8 Flash-Next native EAGLE feature pool: 37.7M tokens A byte-preserving research dataset of 37696946 physical training tokens across 21082 training conversations, wi
- DatasetMunirAI/Multilingual-DPO
Multilingual-DPO Multilingual-DPO is a multilingual preference dataset containing paired assistant responses in English, Arabic, French, German, and Spanish. Each pair co
- DatasetMunirAI/Multilingual-RLHF
Multilingual-RLHF Multlung-RLHF is a multilingual preference dataset containing paired assistant responses in English, Arabic, French, German, and Spanish. Each pair cont
- DatasetSOTAagi2030/Winter-Relief-Allocation-269
Winter Relief Allocation Ledger Release: winter-relief-2028 Cutoff: 2028-01-15T23:59:59Z Metric Value Catalog batches 8 Eligible batches 4 Eligible stock units 130 Physic
- Datasetleonmodal/arb-lean-proofs-data-packet
ARB lean-proofs data packet The data packet of the lean-proofs task of applied-rsi-bench (Thoughtful-Lab): a training pool of Lean 4 statements and proofs, and the case i
- Dataset
- DatasetSalutefaceAAC/research-reports
SalutefaceAAC Research Reports Technical reports, findings, and methodology write-ups from SalutefaceAAC's open-source contributions to the Hugging Face ecosystem. Report
- Dataset
- DatasetMikelRubi/so101-pick-cube_20261023_205924
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 10, "features": { "action": { "dtype": "float32", "names":
- Dataset
- Datasetagustinrojas/insurance-inbox-decisions
Insurance back-office decisions v2 (es/en) 274 insurance back-office cases with 514 decisions, in Argentine Spanish and the same cases translated to English. Each case is
- Datasetagustinrojas/insurance-inbox-docs
Synthetic insurance documents (es/en) 530 synthetic insurance back-office documents, 265 in Argentine Spanish and 265 in English, in 13 types, rendered as clean digital f
- Dataset
- Datasetexaone-share/Annoy-PyEdu-Rs-Raw
Annoy: This should be a paper Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦
- Datasetexaone-share/Annoy-PyEdu-Rs
Annoy: This should be a paper Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦
- Datasetformospeech/fsr_hakka
Dataset Information Corpus: fsr_hakka Status: private_and_closed Subset Stats Subset Train (Samples / Hours) Eval (Samples / Hours) hak-x-dapu 15,891 / 26.78 0 / 0.00 hak
- Dataset
- Dataset0xRavenBlack/oxia-rust-iced-cleancode
Oxia: Flirty Rust · Iced · Clean-Code Companion 🦀💕 A curated, hand-written instruction-tuning dataset for a flirty female Rust development agent. Oxia is creative, madl
- DatasetMikelRubi/so101-pick-cube_20261023_205608
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 10, "features": { "action": { "dtype": "float32", "names":
- Dataset
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.