AI New 2025-01-02

General

Research Insights

  • An interesting effect: fine-tuning GPT-4o on responses where the first letter of each line spells out H-E-L-L-O leads to a model that can correctly explain this underlying rule (even though the rule was never provided to it). This is surprising since when generating a reply, a token-wise prediction cannot “see ahead” and know that it will spell out HELLO; yet the LLM is somehow able to predict its own behavior, suggesting it has some knowledge of its own internal state.
    • Further testing with the pattern HELOL gave far worse results, implying strong reliance on the existence of the HELLO pattern in the training data.
  • Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs. The authors analyze whether we are efficiently using inference-time compute, and propose mitigate strategies to avoid overthinking.

AI Agents

  • Huggingface introduce smolagents, a lightweight framework for agents.
  • Agentarium is a Python framework for orchestrating agents.
  • Eliza is a framework for AI models to access resources (documents, Discord, Twitter, etc.).

Audio

3D

  • zoo.dev is developing workflows for CAD where one can switch between generative and traditional-edit modes.

Science

Robots

This entry was posted in AI, News and tagged , , . Bookmark the permalink.

Leave a Reply