EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
Hacker News·Aug 25·5 min read
Thomson Reuters Launches Its Own Frontier Model
TL;DR
Thomson Reuters has launched Thomson, its proprietary large language model, developed at a significantly lower cost than typical frontier models while maintaining full control. This innovative approach leverages a strong open-source foundation and proprietary data to create a highly specialized AI for professional use.
✦ Why It Matters
Consider exploring Thomson for your professional AI needs to leverage its cost-effective and specialized capabilities.
Key Takeaways
- Explore Thomson's capabilities for legal document review.
- Consider integrating Thomson for specialized AI tasks.
- Evaluate the model's performance against existing tools.
- Stay informed about updates and new features in Thomson.
How It Works
Thomson is built on a strong open-source foundation and proprietary content, utilizing advanced mid-training and post-training techniques. It incorporates insights from subject matter experts throughout its development, ensuring it meets the specific needs of legal professionals.