Large Language Models

Mistral AI and NVIDIA launch Mistral NeMo 12B, a 12-billion-parameter language model for enterprise use, excelling in diverse tasks and easy customisation.

Mistral AI releases Mathstral, a STEM-focused language model built on Mistral 7B. It excels in mathematical reasoning, achieving top performance on MATH and MMLU benchmarks for its size. The model is available for use and fine-tuning.

OpenAI's 'Prover-Verifier Games' method improves AI text clarity. It trains advanced AI to create outputs that simpler AI can easily check, enhancing human understanding.

Codestral Mamba offers linear time inference and infinite sequence handling. It excels in code generation and reasoning, matching top transformer models.

Anthropic's new Developer Console tools, powered by Claude 3.5 Sonnet, help create better AI prompts. Features include prompt generation, testing, and quality grading.

Anthropic launched a programme to fund third-party AI evaluations, focusing on safety and advanced capabilities. The initiative covers areas like cybersecurity, multilingual skills, and societal impacts, aiming to improve AI safety across the industry.