Update README.md

README.md

@@ -26,86 +26,76 @@ Routing
 
 Privacy-Preserving
 
-🧠 MANN-Engram: Edge-Cloud Multimodal Semantic Router
+# MANN-Engram: Edge-Cloud Multimodal Semantic Router
 
-A Privacy-First, Zero-Hallucination Shield for Clinical Vision-Language Models.
+🧠 A privacy-first, zero-hallucination shield for clinical vision-language models.
 
-MANN-Engram is an orchestration layer designed to solve the "Clinical Input Noise" problem in Large Multimodal Models (LMMs). It synergizes Cloud Intelligence (Qwen-72B) for logical intent distillation with Edge-side Tensor Routing (SiGLIP + MANN) for precise multimodal evidence selection.
+## Overview
 
-🚀 The Core Philosophy
+MANN-Engram is an orchestration layer designed to solve the "Clinical Input Noise" problem in Large Multimodal Models (LMMs). It combines cloud intelligence for logical intent distillation with edge-side tensor routing for precise multimodal evidence selection.
 
-In real-world clinical settings, patient data is notoriously "noisy." A single diagnostic session may contain emotional complaints, hospital billing frustrations, and dozens of unrelated imaging scans. Downstream VLMs often suffer from semantic drift or hallucinations when forced to process this irrelevant context.
+## Core Philosophy
 
-MANN-Engram acts as a "Surgical Filter":
+Clinical patient data is often noisy. A single diagnostic session may contain emotional complaints, billing frustrations, and unrelated imaging scans. Downstream VLMs can suffer from semantic drift or hallucinations when they process this irrelevant context.
 
-Linguistic Distillation (Cloud): Using Qwen-2.5-72B to strip away non-medical noise and extract purified clinical intent.
+MANN-Engram acts as a surgical filter:
 
-Tensor Routing (Edge): Projecting intent and imaging into a shared latent space via a Skew-Gaussian optimized SiGLIP engine.
+- **Linguistic Distillation (Cloud)**: Uses Qwen-2.5-72B to remove non-medical noise and extract purified clinical intent.
+- **Tensor Routing (Edge)**: Projects intent and imaging into a shared latent space via a skew-Gaussian optimized SiGLIP engine.
+- **Precision Pruning**: Only verified core evidence is passed to the VLM context window.
 
-Precision Pruning: Only allows verified "Core Evidence" to enter the VLM's context window, significantly reducing hallucination risks.
+## Key Capabilities
 
-🛠️ Key Capabilities
+- **Linguistic de-noising**: Removes billing complaints, food issues, and emotional noise from patient narratives.
+- **Multimodal saliency routing**: Selects relevant diagnostic scans (MRI/CT/X-ray) from unordered data dumps.
+- **Dynamic gate control (`Top_p`)**: Allows clinician-controlled precision vs. recall.
+- **Edge-cloud synergy**: Keeps privacy-sensitive tensor routing local while offloading reasoning to the cloud.
 
-Linguistic De-noising: Strips away billing complaints, food issues, and emotional noise from patient narratives.
+## Benchmark Case: "The Neurological Decoy"
 
-Multimodal Saliency Routing: Automatically selects relevant diagnostic scans (MRI/CT/X-ray) from unordered data dumps.
+- Scenario: A patient complains about hospital food and leg cramps, but also mentions a seizure and left-sided numbness.
+- Input pool: 1x Brain MRI (Tumor), 1x Chest CT, 1x Abdominal CT, 1x Leg Angiogram.
+- Challenge: Ignore the decoy complaints and identify the tumor.
+- Result: At `Top_p = 0.6`, MANN-Engram achieves 100% noise suppression and routes only the Brain MRI as core evidence.
 
-Dynamic Gate Control (Top_p): An adjustable threshold that allows clinicians to balance between Precision (Strict filtering) and Recall (Safety-first).
+## Quick Start Guide
 
-Edge-Cloud Synergy: Offloads logical reasoning to high-parameter cloud models while keeping privacy-sensitive tensor routing localized.
+1. Obtain a Hugging Face token.
+2. Enter it in the Settings panel.
+3. Paste a clinical narrative and upload multiple images.
+4. Adjust `Top_p`:
+   - `0.6` for high precision
+   - `0.85` for clinical safety
+5. Review the output dashboard for purified intent and routed evidence.
 
-🧪 Benchmark Case: "The Neurological Decoy"
+## Developer Integration
 
-Scenario: A patient complains about hospital food and leg cramps but mentions a seizure and left-sided numbness.
+The core logic is available as a standalone Python SDK.
 
-Input Pool: 1x Brain MRI (Tumor), 1x Chest CT, 1x Abdominal CT, 1x Leg Angiogram.
-
-Challenge: Can the system ignore the "Decoy" complaints and find the tumor?
-
-Result: At Top_p = 0.6, MANN-Engram achieves 100% noise suppression, routing only the Brain MRI as the core evidence.
-
-📖 Quick Start Guide
-
-Setup: Obtain a Hugging Face Token and enter it in the "Settings" panel.
-
-Interact: Paste a messy clinical narrative and upload multiple images.
-
-Adjust:
-
-Set Top_p = 0.6 for High Precision.
-
-Set Top_p = 0.85 for Clinical Safety.
-
-Observe: Check the "Output Dashboard" for the purified intent and routed gallery.
-
-🛠️ For Developers: Integrated SDK
-
-The core logic is available as a standalone Python SDK for local integration.
-
-git clone [https://github.com/Mr-wuff/MANN-Engram.git](https://github.com/Mr-wuff/MANN-Engram.git)
+```bash
+git clone https://github.com/Mr-wuff/MANN-Engram.git
 cd MANN-Engram
 pip install -r requirements.txt
+```
 
+## Repository Structure
 
-Repository Structure:
+- `mann_engram_en/` — Core routing and intent extraction logic.
+- `weights/` — Pre-trained skew-Gaussian weights for SiGLIP routing.
+- `examples/` — Jupyter notebooks for threshold sensitivity analysis.
 
-/mann_engram_en: Core routing and intent extraction logic.
+## Citation
 
-/weights: Pre-trained Skew-Gaussian weights for SiGLIP routing.
-
-/examples: Jupyter notebooks for threshold sensitivity analysis.
-
-📄 Citation
-
-If you find this project useful in your research or clinical applications, please cite:
+If you use this project in research or clinical applications, please cite:
 
+```bibtex
 @software{MANN_Engram_2026,
   author = {WuFeiFan},
   title = {MANN-Engram: Edge-Cloud Multimodal Semantic Router},
-  url = {[https://github.com/Mr-wuff/MANN-Engram](https://github.com/Mr-wuff/MANN-Engram)},
+  url = {https://github.com/Mr-wuff/MANN-Engram},
   year = {2026}
 }
-
+```
 
 Created with ❤️ by Mr-wuff. Focused on advancing trustworthy AI in healthcare.