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.
OpenAI unveils its next-generation model **GPT-4o**, expanding advanced features even to free ChatGPT users.
Hugging Face upgraded Transformers Agents into an independent library called smolagents and released it.
The Qwen team has released a large-scale model, **Qwen-Max-0428**, which surpasses the performance of the existing Qwen1.5-110B-Chat.
The Qwen team has released Qwen1.5-110B, a 100B-scale model with performance on par with Llama-3-70B.
OpenAI has released the GPT-3.5 Turbo and Whisper models as APIs for developers.
This proposes an instruction hierarchy that trains LLMs to recognize the priority of system prompts, improving defense against prompt injection attacks.
RAG technology is expected to establish itself as a core technology in the AI ecosystem, beyond being a temporary trend.
Meta has unveiled Llama 3, a new open LLM available in 8B and 70B parameter sizes.
Mistral AI has released Mixtral 8x22B, a new open model that maximizes efficiency and performance, under the Apache 2.0 license.
Qwen has released CodeQwen1.5-7B, an open-source code-specialized LLM trained on 3 trillion code tokens.
The LiveCodeBench leaderboard has launched, addressing data contamination issues and evaluating code generation ability from multiple angles.
OpenAI has established its first office in Asia in Tokyo and unveiled a custom GPT-4 model optimized for Japanese.
OpenAI is improving fine-tuning API features and expanding support for custom models through its custom models program.
It introduces Text-to-SQL technology that combines the DuckDB-NSQL-7B model with the Hugging Face API to query data using natural language.
Enterprise AI reduces the hallucinations and security issues of general-purpose LLMs through RAG and task-specific models.
Replit is building a Replit-native LLM that interacts directly with the IDE environment and automatically fixes code errors.
Qwen has released the 32B-scale Qwen1.5 model, which offers the optimal balance of performance and efficiency.
Harvey partnered with OpenAI to develop a custom model for legal professionals trained on vast amounts of case law data.
AI21 presented 10 use cases for enterprise generative AI using TSMs.
Qwen1.5-MoE-A2.7B achieves performance on par with 7B models while reducing activated parameters to 1/3.
JetBrains is unveiling AI Assistant, which boosts developer productivity by leveraging the OpenAI API, and it is recording the fastest growth ever.
This study demonstrates that stock returns can be predicted by using ChatGPT to analyze the sentiment of news headlines.
Salesforce integrates OpenAI's LLM into the Einstein 1 Platform to deliver enterprise-grade generative AI services with security and trust.
xAI has released the weights and architecture of **Grok-1**, a MoE model with 314 billion parameters, under the Apache 2.0 license.