Spaces:

---

gaia-benchmark

/

leaderboard

like 608

Running on CPU Upgrade

Hello GAIA leaderboard maintainers,

I would like to request permission to submit results to the GAIA leaderboard. Below is the information required for the application:

• Hugging Face Username: iberlins@gmail.com

• Agent/Model Name: diqidiqi

• Brief Description:
  I am working on evaluating and improving an agentic workflow for GAIA-style tasks, focusing on tool-use reliability, multi-step reasoning stability, and reproducible evaluation settings. The primary goal is to benchmark iterative changes (prompting + tool routing + memory strategy) against GAIA to ensure consistent progress and to compare variations under the same evaluation protocol. I would like to submit scores to track these improvements and share reproducible results with the community.

• Optional Links (public):

  • Hugging Face Space demo (optional): https://huggingface.co/insberlins

If you need any additional details (e.g., evaluation configuration, model card, or reproducibility notes), I’m happy to provide them. Thank you for your time and for maintaining the leaderboard.

Best regards,