Upload README.md

README.md

@@ -0,0 +1,39 @@
+# MLE Competitive Model
+
+A high-dimensional sparse distributed memory system with energy-based dynamics and knowledge-based question-answering capabilities.
+
+## Architecture
+
+- **Vector size**: 4096 bits (sparse, ~5% active)
+- **Memory**: Sparse Address Table with online learning
+- **Inference**: Energy minimization via simulated annealing
+- **Knowledge**: 283 facts, 334 words, 53 sequences
+
+## Benchmarks
+
+| Task | Accuracy |
+|------|----------|
+| Question Answering | 92.9% |
+| Analogy Solving | 100.0% |
+| Sequence Completion | 93.3% |
+| Word Retrieval | 100.0% |
+| Generalization | 100.0% |
+| **Overall** | **97.2%** |
+
+## Files
+
+- `mle_best_facts.json`: Knowledge base facts
+- `mle_best_model.npz`: Word vectors and embeddings
+- `mle_best_results.json`: Benchmark results
+- `mle_best_config.json`: Model configuration
+
+## Usage
+
+```python
+import numpy as np
+
+# Load model
+data = np.load("mle_best_model.npz", allow_pickle=True)
+words = data['words']
+vectors = data['vectors']
+```