AI Briefing#llm
The latest AI and developer news about #llm, with the original source and a short summary.
The latest AI and developer news about #llm, with the original source and a short summary.
The Portuguese government has released Amalia, a 9B-parameter open-source LLM.
It's a local tool that improves LLMs' video understanding by detecting scene changes and extracting only key frames.
Proposes a risk control framework that optimizes token usage during LLM inference while keeping the error rate below a certain level.
Apple researchers have proposed the Ctrl-R framework to help LLMs effectively learn the diverse reasoning patterns needed for complex problem solving.
MemoryLLM proposes separating the FFN from Self-Attention in Transformers and using it as a context-free, token-level neural retrieval memory.
This study examines the synergy effects of multi-agent systems that collaborate autonomously without a fixed workflow.
High-quality human annotations were leveraged to train LLMs to perform information filtering and judgment at the level of financial experts.
PorTAL is a base-agnostic technology that allows task adaptation learned once to be easily transferred to a new LLM.
OpenWiki is an open-source tool that automatically generates and maintains documentation so coding agents can better understand codebases.
It covers the architecture and technical considerations for integrating a natural voice interface into existing text-based chatbots.
POSCO DX defined senior engineers' experiential tacit knowledge as UDK that AI has not learned, and verified a method to structure it into a knowledge graph to implement design automation.
Anthropic has released Claude Sonnet 5, which significantly enhances agentic execution capabilities.
OpenAI researchers used GPT 5.5 Pro and Claude Opus 4.8 to solve 9 difficult open problems in mathematics and theoretical computer science.
RadixArk has released Miles, an open-source framework that integrates SGLang, Megatron-LM, and Ray for large-scale LLM reinforcement learning (RL) post-training.
Meituan has released LongCat-2.0, a 1.6-trillion-parameter MoE model, in API form, strengthening agentic coding and long-context processing capabilities.
Ornith-1.0, based on Gemma 4 and Qwen 3.5, has been released as a self-improving open-source model for agentic coding.
Meta has unveiled Brain2Qwerty v2, which combines non-invasive brainwave measurement (MEG) with LLMs to improve real-time sentence decoding performance.
Mistral AI released Leanstral 1.5, a Lean 4 model optimized for automated theorem proving and autoformalization.
Anthropic has released Claude Sonnet 5, an agent-specialized model with significantly improved reasoning, tool use, and coding capabilities.
The PageStorm research model, designed for book-scale creative writing, has been released.
Huawei has open-sourced the OpenPangu-2.0-Flash model, which supports a 512K context length.
FuriosaAI has dramatically accelerated model support speed with SDK 2026.3, which introduces the new kernel framework TCL.
Anthropic's Claude Sonnet 5 model has launched on Vercel AI Gateway, offering improved coding and agentic performance.
This explains three approaches and architectural decision factors for efficiently integrating and scaling ElevenLabs' Text to Speech API.