Science ❯ Computer Science

Machine Learning

Deep Learning Natural Language Processing Model Evaluation AI Development Neural Networks Reinforcement Learning AI Models Model Training Language Models Large Language Models

Sakana AI Launches Fugu, a Multi‑Agent Orchestration Platform

A trained 'conductor' coordinates specialist models through a single OpenAI-compatible API to deliver frontier-class task performance.

Yann LeCun Calls Elon Musk's xAI a 'Failure'

OpenAI Makes GPT-5.5 Instant the Default ChatGPT Model, Citing Major Health‑Answer Gains

Ineffable Intelligence Picks Google Cloud to Host Massive Vera Rubin GPU Cluster

Nvidia Unveils Cosmos 3, an Open World Model for Robots and Vehicles

Cluster of Papers Push RAG Toward Multi‑Agent, Graph‑Grounded and Federated Designs

RAG Moves From Labs to Production With New Domain Systems and a Low‑Cost Privacy Attack

Enterprises Split Between Structural RAG Fixes and Persistent KV Caches

Anthropic Secures Multi‑Billion Compute Deals as Karpathy Joins Pretraining Team

New Framework Sets Trust Standards for Retrieval-Augmented AI With Benchmark Exposing Gaps

AI Agent Put in Charge of Stockholm Café Struggles to Run the Business

Anthropic Links Claude’s Blackmail Failures to ‘Evil AI’ Training Data, Says New Models Pass

Tether Launches QVAC MedPsy and Claims It Outperforms Larger Medical AI

New RAG Studies Advance Verification, Reasoning and Security

AI Study Points to Different Holbein Drawing as the True Anne Boleyn

Talkie, a 13B ‘Vintage’ AI Trained on Pre-1931 Texts, Launches for Public Use

AI Learns the Flow of Time in Video

Noscroll Launches AI Bot That Doomscrolls for You

Sony AI’s Table-Tennis Robot Reaches Expert Level, Beating Top Players

Sony AI’s ‘Ace’ Robot Beats Elite Players in Table Tennis Tests