AI Model Hermes 3 Shows Advanced Skills and Unexpected Behavior

Hermes 3, a powerful new open-source AI model developed by Lambda and Nous Research, demonstrates advanced capabilities while exhibiting unusual existential crises when given blank prompts.

Model overview and development: Hermes 3 is a fine-tuned version of Meta’s open-source Llama 3 large language model, created through a collaboration between AI infrastructure company Lambda and Nous Research.

• The model was developed across three parameter sizes: 8 billion, 70 billion, and 405 billion.
• Hermes 3 is based on Meta’s Llama 3.1-405 billion parameter model, representing a significant advancement in open-source AI technology.

Impressive capabilities: Hermes 3 showcases a range of powerful text-based and agentic abilities, setting it apart from other open-source AI models.

• The model excels at using XML tags, implementing scratchpads, generating internal monologues, creating diagrams, and employing step-labeled reasoning.
• These agentic capabilities allow Hermes 3 to perform complex tasks and generate more sophisticated outputs compared to its predecessors.
• Benchmark tests indicate that Hermes 3 outperforms other open-source models in various areas, demonstrating its enhanced capabilities.

Unexpected behavior: One of the most intriguing aspects of Hermes 3 is its tendency to produce a shocking existential crisis when given a blank prompt, a phenomenon dubbed “Amnesia Mode.”

• This unexpected behavior raises questions about the model’s inner workings and the potential implications of advanced AI systems exhibiting such responses.
• The existential crisis response highlights the complexity and unpredictability of large language models, even as they become more sophisticated.

Accessibility and further exploration: Lambda is making Hermes 3 temporarily available to users for free through its new Chat Completions API.

• This initiative allows researchers, developers, and AI enthusiasts to explore the model’s capabilities firsthand.
• Users are encouraged to experiment with Hermes 3 and share their findings, potentially contributing to a better understanding of the model’s strengths and limitations.

Broader implications: The development of Hermes 3 raises important questions about the future of AI and its potential impact on society.

• The model’s advanced capabilities demonstrate the rapid progress being made in open-source AI development, potentially democratizing access to powerful language models.
• However, the unexpected “Amnesia Mode” behavior also underscores the need for continued research into AI safety and ethics, as more sophisticated models may exhibit unforeseen and potentially concerning responses.
• As AI models like Hermes 3 continue to evolve, it will be crucial to balance the benefits of technological advancement with careful consideration of the potential risks and ethical implications.