This repository contains experimental models designed strictly for academic evaluation and research purposes.

Critical Constraints:

• No Production Deployment: Experimental models must not be deployed in commercial, enterprise, or mission-critical environments under any circumstances.
• No Liability: Experimental models are provided "as-is" without warranties of any kind. The developers assume zero liability for downstream consequences, system integration failures, or regulatory non-compliance resulting from unauthorized deployment.

Hi there 👋

Welcome to ethicalabs.ai — a EU AI research lab based in Italy, focused on sovereign and privacy-preserving AI.

🌟 What We're Building 🇮🇹

Echo-DSRN

A surprise-gated dual-state recurrent architecture (decoder-only) available in 4 families:

• Echo-DSRN, Recurrent Hybrid, 114M parameters.
• Echo-DSRN Hybrid, Transformers Backbone (Qwen2.5) + Echo-DSRN memory injectors, 0.7B and 2.0B parameters.
• Echo-DSRN Classifier, Recurrent Hybrid + Classification Head, 98M parameters.
• Echo-DSRN Embedding, Echo-DSRN-v0.1.3-Embed-Exp (Experimental)

OpenAIRE AI Hackathon 2026 🇪🇺

OpenAIRE Graph Evaluator: Research paper classifier and LLM-as-Judge pipeline

An automated intent classifier and self-improving pipeline designed to turn noisy OpenAIRE graph metadata into clean, trustworthy training data.

• LLM-as-Judge Jury: Asynchronously evaluates papers across 6 research intent categories (methodology, dataset, review, applied, theoretical, and unclassifiable) using a consensus panel of open-weight models.
• Practical Sovereign AI: Powered by our locally-run Echo-DSRN CPU inference, eliminating the need for expensive cloud GPUs.
• Continuous Self-Improvement: Outputs structured JSON verdicts and rationale, feeding verified edge-cases back into the dataset to fine-tune future model iterations.

▶️ Watch the Video

🔬 Research Philosophy

Edge, bare-metal and cloud-native deployment. Intelligence at the edge — bounded by physics, not APIs.

🌈 Get Involved

• GitHub: ethicalabs-ai
• Support: Become a sponsor to ethicalabs.ai

© 2026 Ethicalabs.ai 🇮🇹 — incubated by Dorvan Srl - Operational headquarters: Via Privata Farnese 1/3, 20146 Milano (MI), Italy - Registered address: Via Enrico Cernuschi 4, 20129 Milano (MI), Italy