# Bot18

- **Event:** [Gemini 4 Argon SF Hackathon](https://cerebralvalley.ai/e/gemini-4-argon-sf)
- **When:** Sun, Oct 25 at 9:00 AM – 10:00 PM (PDT)
- **Where:** San Francisco, CA
- **Team:** [shiran wang](https://cerebralvalley.ai/u/ran), [Richard Yi](https://cerebralvalley.ai/u/StingrayYee)
- **GitHub:** https://github.com/blueif16/arg_twin
- **Demo video:** https://youtu.be/0fqARBwj4HU
- **Gallery:** https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/126

Argon turns real robot footage into a simulated world shaped around the task, then teaches itself the task inside that world. With practice, work moves from slow reasoning into fast routines, so each attempt is more reliable and cheaper. The "policy" is Argon plus its harness: tools, compiled skills and a watcher.

## More from Gemini 4 Argon SF Hackathon

- [Two Lenses](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/1)
- [Business C Suite](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/123)
- [Agent Gym](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/124)
- [Argon Music](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/125)
- [Heirloom](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/127)
- [Infinite Peaks](https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/128)

---

Markdown version of https://cerebralvalley.ai/e/gemini-4-argon-sf/hackathon/gallery/126. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt