Latest stories
Google DeepMind Unveils WeatherNext 3: AI Model with Real-Time Satellite Data and 5x Sharper Forecasts
Google DeepMind and Google Research introduce WeatherNext 3, an AI model that uses real-time satellite data to deliver hourly, hig...
OpenAI Launches $1B 'Daybreak' Initiative to Protect Essential Services
OpenAI announces a $1 billion commitment under the 'Daybreak for Frontline Defenders' program to expand access to frontier cyber A...
OpenAI's GPT-6 Astra Helps Legora Review 41 Documents in Minutes, Boosts Performance by 40%
Legora, a financial review platform, used OpenAI's GPT-6 Astra to review 41 documents in minutes, catching all four planted errors...
Playco Cuts Manual Fixes by 50% Prototyping Games with GPT-6 Astra
OpenAI reports that Playco, using GPT-6 Astra, reduced manual fixes by 50% while building three game prototypes from a single grey...
Qwen and Taobao Launch E-Commerce Bench to Test Long-Horizon AI Store Management
Alibaba's Qwen team, in partnership with Taobao & Tmall Group, releases E-Commerce Bench, a realistic simulation benchmark that ev...
Alibaba's Qwen-Drive-1.0: A Unified Vision-Language Model for Autonomous Driving
Qwen-Drive-1.0, built on Qwen3.5-4B, unifies 3D perception, visual QA, and motion planning in a single VLM without altering its ar...
NVIDIA Nemotron Models Outperform Top Humans at IOI 2026 Coding Competition
A new post-training pipeline from NVIDIA, combining curated problems, synthetic reasoning, SFT, and RL, enables Nemotron models to...
New Distillation Method Tackles Diversity Bottleneck in Reasoning Models
Researchers propose IDA-OPD, a distillation technique that improves diversity transfer in reasoning models by selectively preservi...
Cliff: New Reward Shaping Method Boosts LLM Reasoning by Detecting First Mistakes
Researchers propose Cliff, a reward shaping strategy that uses an off-the-shelf LLM to detect the first reasoning error, improving...
Hugging Face Unveils VibeVoice-ASR-Streaming: Real-Time Speaker-Attributed Speech Recognition
Researchers present VibeVoice-ASR-Streaming, an LLM-based end-to-end model that unifies streaming speech recognition and speaker d...
SimLoss: Single-Pass Image Captioning Matches Multi-Stage Quality at 20x Lower Latency
A new Hugging Face paper introduces SimLoss, an embedding-space contrastive objective that enables single-pass fine-grained image...
ZipTok3D: High-Fidelity 3D Reconstruction from Ultra-Short Token Sequences
New tokenizer organizes 3D geometry into compact global-token prefixes, achieving reconstruction quality comparable to 32-token ba...