Spaces:
Sleeping
Sleeping
Update README.md
Browse files
README.md
CHANGED
|
@@ -17,10 +17,12 @@ tags:
|
|
| 17 |
|
| 18 |
# ExecAssist — Executive Assistant Environment
|
| 19 |
|
| 20 |
-
|
|
|
|
|
|
|
| 21 |
|
| 22 |
**Live environment:** https://devanshudon-exec-assist.hf.space
|
| 23 |
-
**Mini-blog:**
|
| 24 |
**Training notebook:** [`train_colab.ipynb`](./train_colab.ipynb)
|
| 25 |
|
| 26 |
---
|
|
|
|
| 17 |
|
| 18 |
# ExecAssist — Executive Assistant Environment
|
| 19 |
|
| 20 |
+
> **A 0.5B-parameter model, trained for 90 minutes on a free Colab T4, beats an untuned 120B-parameter frontier model on this environment by 2.4×.** Built for the OpenEnv Hackathon (Apr 2026), Theme #3.2 — Personalized Tasks.
|
| 21 |
+
|
| 22 |
+
An OpenEnv environment where AI agents learn to manage email and calendar like a human executive assistant — read incoming requests, write professional replies, find calendar slots that don't clash, propose alternatives when they do. Three tasks at increasing difficulty, three independent reward graders, and four anti-reward-hacking penalties that we have direct evidence of catching the model in the act.
|
| 23 |
|
| 24 |
**Live environment:** https://devanshudon-exec-assist.hf.space
|
| 25 |
+
**Mini-blog:** _(link will go here once published)_
|
| 26 |
**Training notebook:** [`train_colab.ipynb`](./train_colab.ipynb)
|
| 27 |
|
| 28 |
---
|