Zhipu AI Releases ‘Glyph’: An AI Framework for Scaling the Context Size by way of Visible-Textual content Compression
Can we render lengthy texts as pictures and use a VLM to attain 3–4× token compression, preserving accuracy whereas scaling ...
Tag: Releases
OpenAI Releases Shared Undertaking Function To All Customers
OPenAI introduced that it's upgrading all ChatGPT accounts to be eligible for the challenge sharing characteristic, which allows customers to ...
Kong Releases Volcano: A TypeScript, MCP-native SDK for Constructing Manufacturing Prepared AI Brokers with LLM Reasoning and Actual-World actions
Kong has open-sourced Volcano, a TypeScript SDK that composes multi-step agent workflows throughout a number of LLM suppliers with native ...
Baidu’s PaddlePaddle Staff Releases PaddleOCR-VL (0.9B): a NaViT-style + ERNIE-4.5-0.3B VLM Focusing on Finish-to-Finish Multilingual Doc Parsing
How do you exchange advanced, multilingual paperwork—dense layouts, small scripts, formulation, charts, and handwriting—into trustworthy structured Markdown/JSON with state-of-the-art accuracy ...
Anthropic AI Releases Petri: An Open-Supply Framework for Automated Auditing by Utilizing AI Brokers to Check the Behaviors of Goal Fashions on Various Eventualities
How do you audit frontier LLMs for misaligned conduct in life like multi-turn, tool-use settings—at scale and past coarse mixture ...
H Firm Releases Holo1.5: An Open-Weight Laptop-Use VLMs Centered on GUI Localization and UI-VQA
H Firm (A french AI startup) releases Holo1.5, a household of open basis imaginative and prescient fashions purpose-built for computer-use ...
Google AI Releases VaultGemma: The Largest and Most Succesful Open Mannequin (1B-parameters) Educated from Scratch with Differential Privateness
Google AI Analysis and DeepMind have launched VaultGemma 1B, the biggest open-weight massive language mannequin educated totally with differential privateness ...
Google AI Releases EmbeddingGemma: A 308M Parameter On-System Embedding Mannequin with State-of-the-Artwork MTEB Outcomes
EmbeddingGemma is Google’s new open textual content embedding mannequin optimized for on-device AI, designed to stability effectivity with state-of-the-art retrieval ...
StepFun AI Releases Step-Audio 2 Mini: An Open-Supply 8B Speech-to-Speech AI Mannequin that Surpasses GPT-4o-Audio
The StepFun AI staff has launched Step-Audio 2 Mini, an 8B parameter speech-to-speech massive audio language mannequin (LALM) that delivers ...
NVIDIA AI Releases ProRLv2: Advancing Reasoning in Language Fashions with Prolonged Reinforcement Studying RL
What Is ProRLv2? ProRLv2 is the newest model of NVIDIA’s Extended Reinforcement Studying (ProRL), designed particularly to push the boundaries ...
Trending.
