Update README.md

README.md

@@ -15,16 +15,17 @@ AnveshAI Edge is a terminal-based AI assistant designed to run entirely on-devic
 ## Table of Contents
 
 1. [Architecture Overview](#architecture-overview)
-2. [Components](#components)
-3. [Advanced Math Engine](#advanced-math-engine)
-4. [Reasoning Engine](#reasoning-engine)
-5. [Getting Started](#getting-started)
-6. [Usage Examples](#usage-examples)
-7. [Commands](#commands)
-8. [Design Principles](#design-principles)
-9. [File Structure](#file-structure)
-10. [Technical Details](#technical-details)
-11. [Links](#link)
+2. [Benchmark test & Evaluation experiments](#Benchmark-test-&-Evaluation-experiments)
+3. [Components](#components)
+4. [Advanced Math Engine](#advanced-math-engine)
+5. [Reasoning Engine](#reasoning-engine)
+6. [Getting Started](#getting-started)
+7. [Usage Examples](#usage-examples)
+8. [Commands](#commands)
+9. [Design Principles](#design-principles)
+10. [File Structure](#file-structure)
+11. [Technical Details](#technical-details)
+12. [Links](#link)
 
 ---
 
@@ -64,6 +65,12 @@ User Input
                                         └► LLM with CoT context → User
 ```
 
+## Benchmark test & Evaluation experiments
+
+<img src="https://raw.githubusercontent.com/AnveshAI/AnveshAI-Edge/refs/heads/main/diagram/download%20(1).png">
+<img src="https://raw.githubusercontent.com/AnveshAI/AnveshAI-Edge/refs/heads/main/diagram/download.png">
+
+
 ### Key Design Principle
 
 > **Correctness-first:** For mathematics, the symbolic engine computes the exact answer *before* the LLM is called. The LLM's only task is to explain the working — it cannot invent a wrong answer.