Spaces:
Running
Running
Create AxiomSet Labs organization card
Browse files
README.md
CHANGED
|
@@ -1,10 +1,50 @@
|
|
| 1 |
---
|
| 2 |
-
title:
|
| 3 |
-
emoji:
|
| 4 |
-
colorFrom:
|
| 5 |
-
colorTo:
|
| 6 |
sdk: static
|
|
|
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
title: AxiomSet Labs
|
| 3 |
+
emoji: 📚
|
| 4 |
+
colorFrom: blue
|
| 5 |
+
colorTo: green
|
| 6 |
sdk: static
|
| 7 |
+
short_description: Expert STEM data and evaluation for advanced AI
|
| 8 |
pinned: false
|
| 9 |
---
|
| 10 |
|
| 11 |
+
# AxiomSet Labs
|
| 12 |
+
|
| 13 |
+
**Research-grade STEM data for AI systems that need to reason.**
|
| 14 |
+
|
| 15 |
+
AxiomSet Labs designs expert-authored training, post-training, benchmark, and evaluation data for advanced AI systems.
|
| 16 |
+
|
| 17 |
+
## What we build
|
| 18 |
+
|
| 19 |
+
- **Expert STEM datasets** — difficult, domain-specific tasks for training and post-training
|
| 20 |
+
- **Benchmarks and evaluations** — original evaluation sets, scoring rubrics, and error analysis
|
| 21 |
+
- **Scientific code and reasoning data** — executable tasks that combine technical reasoning with computation
|
| 22 |
+
- **Dataset quality and verification** — expert review, consistency checks, provenance, documentation, and versioning
|
| 23 |
+
|
| 24 |
+
## Domains
|
| 25 |
+
|
| 26 |
+
Mathematics · Physics · Chemistry · Materials Science · Biology · Scientific Computing
|
| 27 |
+
|
| 28 |
+
## Our approach
|
| 29 |
+
|
| 30 |
+
1. Define the model capability and acceptance criteria.
|
| 31 |
+
2. Author tasks with relevant subject-matter expertise.
|
| 32 |
+
3. Review solutions, code, and test behavior.
|
| 33 |
+
4. Validate structure, metadata, and release readiness.
|
| 34 |
+
5. Deliver documented, versioned datasets and evaluation assets.
|
| 35 |
+
|
| 36 |
+
## Current work
|
| 37 |
+
|
| 38 |
+
Our first five-domain scientific-code sample contains 30 tasks across 30 distinct subdomains. It is undergoing final rights, licensing, and private-release checks before publication.
|
| 39 |
+
|
| 40 |
+
## Work with us
|
| 41 |
+
|
| 42 |
+
We support AI labs, model teams, research organizations, and data partners that need challenging, expert-verified STEM data.
|
| 43 |
+
|
| 44 |
+
- Website: [axiomsetlabs.com](https://axiomsetlabs.com/)
|
| 45 |
+
- Email: [hello@axiomsetlabs.com](mailto:hello@axiomsetlabs.com)
|
| 46 |
+
- LinkedIn: [AxiomSet Labs](https://www.linkedin.com/company/axiomset-labs)
|
| 47 |
+
|
| 48 |
+
---
|
| 49 |
+
|
| 50 |
+
*AxiomSet Labs is an independent AI data and evaluation company.*
|