Broaden the organisation card to AI for science

README.md

@@ -7,44 +7,74 @@ sdk: static
 pinned: false
 ---
 
-<img src="https://huggingface.co/spaces/mycel-ai/README/resolve/main/mycel-banner.png" alt="Mycel: open models for research, built in Germany" width="100%">
+<img src="https://huggingface.co/spaces/mycel-ai/README/resolve/main/mycel-banner.png" alt="Mycel: AI for science and research, built in Germany" width="100%">
 
-**Mycel** is an AI-native company from Germany. We build AI solutions for
-research institutions, enterprises and cities, from strategy and compliance
-to autonomous agents. They are designed for the EU AI Act, GDPR and
-on-premise deployment.
+**Mycel** builds AI for science, research and innovation. We develop
+specialised models that reason over evidence, coordinate tools and support the
+work behind discovery, with people in control.
 
-On Hugging Face, we publish **open-weight models that run on hardware you
-own**. Each release comes with its full evaluation, its limitations stated
-plainly, and a DOI so you can cite it.
+Like mycelia networks that connect a forest underground, we build systems that
+link what matters. Mycel is rooted in Germany, and its work spans research
+institutions, enterprises and cities.
 
-## Featured model
+**How we release models**
+
+- **Open weights.** They run on hardware you own.
+- **Full evaluations.** Each release documents how it was tested.
+- **Plain limitations.** We say what a model cannot do.
+- **Citable.** Every release gets a DOI.
+
+## Models
+
+### Numo-1-E2B (released)
 
 <a href="https://huggingface.co/mycel-ai/Numo-1-E2B"><img src="https://huggingface.co/mycel-ai/Numo-1-E2B/resolve/main/assets/numo-1-banner.png" alt="Numo-1-E2B: calibrated scientific-paper decisions on hardware you own" width="100%"></a>
 
-**[Numo-1-E2B](https://huggingface.co/mycel-ai/Numo-1-E2B)** reads a paper's
-title and abstract and decides what kind of study it is, which field it
-belongs to, or whether it is relevant to another paper. It gives a calibrated
-probability for every answer.
+**[Numo-1-E2B](https://huggingface.co/mycel-ai/Numo-1-E2B)** is a fast
+decision model for scientific literature. From a paper's title and abstract,
+it decides what kind of study the paper is, which field it belongs to, or
+whether it is relevant to another paper, with a calibrated probability for
+every answer.
+
+- **Accuracy:** 94.3% on 600 PubMed papers labelled by NLM indexers.
+- **Calibration:** the lowest calibration error of every model we tested.
+- **Speed:** about 124 ms per decision on one RTX 3060, fully offline.
+- **Licence:** Apache-2.0, built on Google's Gemma 4 E2B.
+- **DOI:** [10.57967/hf/11188](https://doi.org/10.57967/hf/11188).
 
-- **Accurate.** 94.3% on 600 PubMed papers labelled by NLM indexers, within
-  1–2 points of the best hosted decision models.
-- **Trustworthy probabilities.** It has the lowest calibration error of every
-  model we tested.
-- **Fast and local.** About 124 ms per decision on one RTX 3060, and fully
-  offline.
-- **Open.** Apache-2.0, built on Google's Gemma 4 E2B.
-- **Citable.** DOI [10.57967/hf/11188](https://doi.org/10.57967/hf/11188).
+### Ark-1 (research in progress)
 
-## Also from Mycel
+**[Ark-1](https://mycel-ai.de/en/ark-1)** is the first model in the Ark series.
+It is a research project developing specialised AI that reasons over
+evidence, coordinates tools and supports research and innovation workflows,
+starting with research funding.
+
+- **Now:** post-training on MareNostrum 5 at the Barcelona Supercomputing
+  Center.
+- **Next:** Ark-1.5 is planned with the HammerHAI AI Factory.
+- **Release:** public releases will follow evaluation and review.
+
+This is ongoing research. It makes no claim of a released model or of
+validated performance. The infrastructure acknowledgements describe computing
+access only and do not imply endorsement.
+
+## Products and initiatives
 
 - **[ResearchArk](https://researchark.eu)** is our agentic operating system
-  for research, innovation and funding. It helps with finding funding,
-  building consortia and writing applications.
-- **Open initiatives:** open frameworks, open-source tools and open-access
-  resources. See [mycel-ai.de](https://mycel-ai.de/en).
+  for research, innovation and funding. It helps you discover funding, develop
+  proposals, build consortia and explore the research graph. It is live and
+  free to start.
+- **[Agentpower](https://mycel-ai.de/en/agentpower)** is an open,
+  public-domain framework. It measures the cognitive work output of human–AI
+  systems.
+- **[More initiatives](https://mycel-ai.de/en/initiatives):** open
+  frameworks, free and open-source tools, and open-access resources.
+
+## Collaborate
 
-## Contact
+We welcome conversations with researchers, institutions and partners working
+on reliable AI for research and innovation.
 
 [mycel-ai.de](https://mycel-ai.de) · [dev@mycel-ai.de](mailto:dev@mycel-ai.de) ·
-Mycel UG (haftungsbeschränkt), Germany
+[LinkedIn](https://www.linkedin.com/company/mycel-ai) · Mycel UG
+(haftungsbeschränkt), Berlin, Germany