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…
Yesterday
- Datasetasd12sadad2/Annoy-PyEdu-Rs
Annoy: This should be a paper Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦
- Dataset
- Dataset
- DatasetRcarvalo/lfm25-agent-230m-train
LFM2.5-Audio 230M agent: training rows (web_search / db_query, French and English) Training set of the 230M LFM2.5-Audio student distilled from lfm2a-fr S12, trained to d
- DatasetCAMeL-Lab/BAREC-LexRank
BAREC-LexRank: A Ranked Arabic Readability Lexicon Summary BAREC-LexRank is a ranked readability lexicon of 40K Arabic open-class lemmas (adjectives, nouns, proper nouns,
- Dataset
- Datasetrahulsharmavishwakarma/vesper-asr-benchmark
Vesper ASR benchmark Real speech-recognition outputs from on-device benchmark runs · Text and IDs, no private audio What's excluded · Getting the audio · Fields · Caveats
- Dataset
- DatasetAnnoor24/power-grid-ai-blindspot
Environmental Blindness in AI Power & Infrastructure Modeling Hi! This repository holds the files for my Fatima Fellowship application. For this project, I tested a major
- Datasetyanguanyue/DFTA-generated-results
DFTA generated images and masks Lossless backup of generated image files, corresponding masks where supplied, and separately labeled comparison illustrations from DFTA ma
- Datasetwoog/alphaxiv-pangram-results
alphaXiv Pangram Results Pangram AI-text detection results for 3,725 arXiv papers, as shown on alphaXiv paper pages, with the paper PDFs and the text Pangram scored. alph
- Datasetkaggle-benchmarks-leads/Annoy-PyEdu-Rs-Raw
Annoy: This should be a paper Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦
- Datasetkaggle-benchmarks-leads/Annoy-PyEdu-Rs
Annoy: This should be a paper Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦
- DatasetWTFO/ru_extra
WTFO/ru_extra 俄語 ASR 補充資料：俄語音訊 → 俄文轉寫。用來補充基線 YuKuanFu/ru（約 4,465.6 h）。 Russian ASR supplement (Russian audio → Russian text) for the YuKuanFu/ru baseline (~4,465.6 h). co
- Datasetlerobot-projects/2_bin_recycle_merged_20261022
This dataset was created using LeRobot. Dataset Structure meta/info.json: { "codebase_version": "v3.0", "fps": 30, "features": { "action": { "dtype": "float32", "names":
- Datasetassomaidi/3_ENG_BENCHMARK_MALAGASY_3_1224
language: - mg - en license: cc-by-4.0 task_categories: - automatic-speech-recognition - text-to-speech - translation size_categories: - 1K<n<10K tags: - audio - speech -
- DatasetHebrish/ASR-Error-List
Hebrish ASR Error List The words the Hebrish speech-to-text fine-tunes still get wrong in real dictation, what they were heard as, and how far targeted retraining on each
- Datasetweili-0234/flex-pd-experiment-traces
flex-pd / sgl-p2p experiment traces (Sept-Oct 2026) Raw experiment traces, request dumps, engine/router logs and simulator outputs from the flex-pd (SGLang prefill/decode
- DatasetAtomicChat/Flint-GGUF-metrics
Flint GGUF metrics Everything behind AtomicChat/Flint-GGUF: the files there are built with our own importance matrix, and every number on its card comes from here. result
- Dataset
- Datasetconnections-dev/_natural_s0_e10_original_1_medium_0.7_16384_gpt-6-astra
Dataset: connections-dev/connection_queries_jan12 This dataset was generated using the inference script with the following configuration: Inference Parameters Model Confi
- Dataset
- Datasetnohaah/laya-home-fine-tune-ko-en-bilingual-training-set
Gantry smart-home commands (English + Korean) Synthetic smart-home commands in English and Korean ("주방 불 켜줘", "현관문 잠가줘", "turn off the kitchen lights") labelled with the
- Dataset
- Datasetnohaah/laya-home-fine-tune-en-training-set
Gantry smart-home commands (English) Synthetic English smart-home commands ("turn off the kitchen lights", "lock the front door", "make it a bit warmer") labelled with th
- Dataset
- Dataset
- DatasetAeliaTR/aelia-stt-v6-veri
Aelia ASAMA-6 STT verisi - 17521 satir / 12 dil AeliaTR STT v6 (Whisper large-v3-turbo + LoRA) egitimi icin tek kaynak. Ses, 16 kHz tek kanal int16 olarak dil basina pake
- Datasetnikhilr01/sen-bridge
SEN-Bridge Paired Sentinel-1 radar and Sentinel-2 optical image crops (128x128, 20 m/pixel) for benchmarking flow-matching and bridge models on radar-to-optical translati
- Dataset
- Dataset
- Dataset
- Dataset
- Datasethamlab-princeton/pastcasting-factual-52k
Pastcasting: controlled factual bank, 1921–1972 The complete frozen 52,000-question factual recognition bank: exactly 1,000 questions per year for 52 years. Each year has
- Datasetasfafaf4546/Annoy-PyEdu-Rs-Raw
Annoy: This should be a paper Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦
- Datasetasfafaf4546/Annoy-PyEdu-Rs
Annoy: This should be a paper Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦
- DatasetNehalNN10/frontier-model-blindspot-eval
Evaluating Western Defaults in Small Open-Weight Models Model tested: Qwen/Qwen2.5-3B-InstructWhat's in here: An evaluation notebook (evaluation_notebook.ipynb) and the r
- Datasetzxcgvbgfdg/Annoy-PyEdu-Rs-Raw
Annoy: This should be a paper Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦
- Datasetzxcgvbgfdg/Annoy-PyEdu-Rs
Annoy: This should be a paper Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦
- Datasetdfsfsdg5657/Annoy-PyEdu-Rs-Raw
Annoy: This should be a paper Title 📑 Paper | 🌐 Project Page | 💾 Released Resources | 📦
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.