GenAI security and governance firm Patronus AI launched Lynx, an open-source hallucination detection model designed to address hallucinations in large language models (LLMs). Patronus AI claims that Lynx outperforms other LLMs such as GPT-4o and Claude-3 in detecting hallucinations across various domains, including finance and medicine.
The company has also open-sourced HaluBench, a benchmark sourced from real-world domains to assess accuracy of LLM responses in domain-specific scenarios.
Patronus AI plans to monetize Lynx through its enterprise solutions such as scalable API access and bespoke integrations suited to specific business needs.
Analyst QuickTake: Lynx’s development utilizes the USD 17 million raised by the company in May 2024 . Specifically, proceeds were directed toward scaling Patronus AI's research, engineering, and sales divisions, as well as training evaluation models and formulating new industry-standard benchmarks.
By using this site, you agree to allow SPEEDA Edge and our partners to use cookies for analytics and personalization. Visit our privacy policy for more information about our data collection practices.