the-decoder3d
Nvidia improves Meta's Llama model with new training approach
Nvidia has introduced a new large language model that outperforms others in alignment benchmarks. The company achieved this through a special training procedure combining evaluation and preference ...
the-decoder3d
Perplexity introduces 'Internal Knowledge Search' for companies
Perplexity, the AI-powered search tool, is rolling out a new "Internal Knowledge Search" feature for its Pro and Enterprise Pro subscribers. This function allows users to search both public web ...
the-decoder4d
Tiny open-source image model Meissonic offers impressive image quality for its size
A new open-source AI model called Meissonic can generate high-quality images using only a billion parameters. This compact size could enable local text-to-image applications, especially on mobile ...
the-decoder4d
Google wants to turn its NotebookLM minute into a ChatGPT moment
Google is positioning its AI-powered NotebookLM as a strong competitor to ChatGPT. The company has added new features and shifted its strategy to help the tool gain traction. NotebookLM has been ...
the-decoder4d
ChatGPT app now available for Windows and macOS
OpenAI has released an early version of the Windows app for ChatGPT. A full version is expected to follow by the end of the year. The app offers the same features known from the ChatGPT web version.
the-decoder8d
'OCR 2.0' model converts images of text, formulas, notes, and shapes into editable text
According to the researchers, an OCR 2.0 model uses a unified end-to-end architecture and requires fewer resources than large language models, while being versatile enough to recognize more than just ...
the-decoder9d
Apple AI researchers question OpenAI's claims about o1's reasoning capabilities
A new study by Apple researchers, including renowned AI scientist Samy Bengio, calls into question the logical capabilities of today's large language models - even OpenAI's new "reasoning model" o1.
the-decoder7d
AI model simulates Counter-Strike with 10 FPS on a single RTX 3090
Researchers have created an AI model that can simulate a janky version of Counter-Strike: Global Offensive (CS:GO) inside a neural network. The model, called DIAMOND (Diffusion for World Modelling), ...
the-decoder4d
Apple's local AI agent framework paves the way for more useful Apple Intelligence
Apple's AI research team has introduced CAMPHOR, a new AI framework designed to process complex user queries locally on mobile devices while preserving user privacy. CAMPHOR (Collaborative Agents for ...
the-decoder4d
Researchers unveil first framework to assess AI models against EU regulations
Researchers from ETH Zurich, INSAIT, and LatticeFlow AI have created the first comprehensive evaluation platform for generative AI models in the context of the EU AI Act. Their findings reveal ...
the-decoder9d
OpenAI introduces experimental multi-agent framework "Swarm"
OpenAI has released a new open-source framework called "Swarm" on GitHub. The company describes it as an experimental tool for creating, orchestrating, and deploying multi-agent systems. Swarm aims to ...
the-decoder9d
Anthropic CEO thinks AGI in 2026 is possible, but says it could take much longer
While acknowledging significant uncertainties, Amodei suggests that extremely capable AI systems, which some might call "AGI," could emerge as early as 2026. He cautions, however, that it could take ...