Update README.md

README.md

@@ -1,39 +1,39 @@
----
-title: NEBULA Photonic Neural Network Demo
-emoji: 🌌
-colorFrom: blue
-colorTo: purple
-sdk: gradio
-sdk_version: 4.0.0
-app_file: app.py
-pinned: false
-license: mit
----
-
-# NEBULA Photonic Neural Network Demo
-
-Interactive demo for **NEBULA** (Neural Emulated Beings Using Light Architecture) - a revolutionary photonic neural network for ARC-AGI spatial reasoning.
-
-## Features
-- Photonic raytracing simulation with authentic light interference
-- Quantum memory neurons (4-qubit superposition states) 
-- Holographic pattern storage using FFT spatial recognition
-- Spatial reasoning transformer for grid analysis
-
-## Performance
-- **25% exact match accuracy** on official ARC-AGI benchmark
-- **97.3% cell-level accuracy** in spatial transformations
-- **100% perfect performance** on rotation and scaling patterns
-
-## Model Details
-- **Architecture**: Photonic + Quantum + Spatial Transformer
-- **Parameters**: 12,852,808
-- **Training Data**: 100% official ARC-AGI from François Chollet
-- **Philosophy**: "Simple solutions for complex problems, without placeholders and with truth first"
-
-## Try NEBULA
-Upload an ARC-like grid pattern and watch NEBULA apply its photonic intelligence to predict transformations!
-
-**Author**: Francisco Angulo de Lafuente - Project NEBULA Team  
-**Model**: [Agnuxo/nebula-photonic-HRM-ARC-2](https://huggingface.co/Agnuxo/nebula-photonic-HRM-ARC-2)  
-**License**: MIT
+---
+title: NEBULA Photonic Neural Network Demo
+emoji: 🌌
+colorFrom: blue
+colorTo: purple
+sdk: gradio
+sdk_version: 5.43.1
+app_file: app.py
+pinned: false
+license: mit
+---
+
+# NEBULA Photonic Neural Network Demo
+
+Interactive demo for **NEBULA** (Neural Emulated Beings Using Light Architecture) - a revolutionary photonic neural network for ARC-AGI spatial reasoning.
+
+## Features
+- Photonic raytracing simulation with authentic light interference
+- Quantum memory neurons (4-qubit superposition states) 
+- Holographic pattern storage using FFT spatial recognition
+- Spatial reasoning transformer for grid analysis
+
+## Performance
+- **25% exact match accuracy** on official ARC-AGI benchmark
+- **97.3% cell-level accuracy** in spatial transformations
+- **100% perfect performance** on rotation and scaling patterns
+
+## Model Details
+- **Architecture**: Photonic + Quantum + Spatial Transformer
+- **Parameters**: 12,852,808
+- **Training Data**: 100% official ARC-AGI from François Chollet
+- **Philosophy**: "Simple solutions for complex problems, without placeholders and with truth first"
+
+## Try NEBULA
+Upload an ARC-like grid pattern and watch NEBULA apply its photonic intelligence to predict transformations!
+
+**Author**: Francisco Angulo de Lafuente - Project NEBULA Team  
+**Model**: [Agnuxo/nebula-photonic-HRM-ARC-2](https://huggingface.co/Agnuxo/nebula-photonic-HRM-ARC-2)  
+**License**: MIT