gfernandf1

5 ms·

1. 1.▲
   by gfernandf1 3mo ago

   https://zenodo.org/records/19438943
2. 2.▲
   by gfernandf1 3mo ago

   https://zenodo.org/records/21108157
3. 3.▲
   Action Preflight: consequence-aware admission for LLM agent actions(github.com)

   1 points by gfernandf1 3mo ago|2 comments
4. 4.▲
   by gfernandf1 5mo ago

   One thing I didn’t make clear in the post: The core issue is that prompt-based agents recompute reasoning every time. ORCA tries to make reasoning reusable — more like software, less like a one-shot prompt. Working on a minimal example to s
5. 5.▲
   by gfernandf1 5mo ago

   you can also download the paper from zenodo https://zenodo.org/records/19438943 slightly different version but the core ideas remain the same
6. 6.▲
   Claude Leak Confirms It: LLM Systems Are Architecture, Not Prompts (Orca)(github.com)

   2 points by gfernandf1 5mo ago|2 comments
7. 7.▲
   by gfernandf1 6mo ago

   Comments welcome!
8. 8.▲
   by gfernandf1 6mo ago

   If helpful, I can also share the PDF directly or specific sections (design principles, execution model, etc.)
9. 9.▲
   Orca: A cognitive runtime layer for agents (open source)(github.com)

   3 points by gfernandf1 6mo ago|2 comments
10. 10.▲
    Orca – moving beyond prompt-based agents(github.com)

    1 points by gfernandf1 6mo ago|0 comments