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.
Qwen has released Qwen2-Math, a series of math-specialized LLMs that maximize mathematical problem-solving ability.
Rakuten is leveraging OpenAI's API and RAG technology to analyze vast amounts of data, driving innovation in customer service and B2B business.
Mistral AI announced model fine-tuning, an alpha version of Agents, and the release of SDK 1.0.
Anthropic has launched Claude Pro, a paid subscription plan that increases usage on Claude.ai by 5x, in the US and UK.
Anthropic has released Claude 3 Haiku, its fastest and most affordable model.
Anthropic unveiled Claude 2.1, featuring a 200K context window and a 2x reduction in hallucination rate.
Anthropic unveiled its new model Claude 2, featuring substantial improvements in performance, safety, and context window.
Google has released Gemma 2 2B for on-device use, the safety classifier ShieldGemma, and the interpretability tool Gemma Scope.
Cohere has released Prompt Tuner in beta, a feature that automatically improves prompts.
Kurly automated review summarization and analysis by calling Gemini Pro directly from BigQuery, without building a separate model.
Mistral AI has unveiled Mistral Large 2, a high-performance model with 123 billion parameters.
Meta's Llama 3.1 model series has been released, offering 128K context and multilingual support.
Introduces a method for efficiently running the Mistral 7B model on Mac by leveraging Apple's new Core ML features.
OpenAI has launched a Structured Outputs feature that ensures model outputs comply 100% with a specified JSON schema.
OpenAI has unveiled **GPT-4o mini**, a small model that is over 60% cheaper than GPT-3.5 Turbo while offering enhanced performance.
Mistral collaborated with NVIDIA to release Mistral NeMo, a 12B model with enhanced Korean language performance.
Prover-Verifier games can simultaneously improve the readability and verifiability of outputs generated by LLMs.
SkillsEngine's product and design team adopted Replit Teams, dramatically speeding up prototyping and internal tool development.
Mistral AI unveiled Codestral Mamba, based on the Mamba architecture capable of linear inference and unlimited sequence handling.
Mistral AI has unveiled **Mathstral**, a STEM-specialized model designed for complex mathematical reasoning.
This explores how to build robust enterprise AI solutions that account for LLM performance and safety, along with future trends.
The NuminaMath 7B TIR model won the first AI Math Olympiad (AIMO) competition.
Transformers models from Hugging Face can now be loaded and used directly in KerasHub.
France's Banque des Territoires built a RAG solution that ensures data sovereignty.