Update README.md

README.md

@@ -103,7 +103,7 @@ The training set is constructed via a multi-model annotation framework:
 | 4 | **Eva-4B (Ours)** | **81.3%** | **0.807** |
 | 5 | GPT-5.2 | 80.5% | 0.805 |
 
-Note: by the accuracy values in the paper’s table, Eva-4B is above GPT-5.2. The paper also states Eva-4B **“ranks 5th overall and 2nd among open-source models (after GLM-4.7)”**, which appears inconsistent with the raw ordering implied by the accuracies.
+Note: based on the accuracy values, Eva-4B is **2nd among open-source models**, after GLM-4.7 (82.6%).
 
 ### Per-class F1 (Eva-4B)
 
@@ -151,7 +151,7 @@ Response format:
 {"reason": "brief explanation under 100 characters", "label": "direct|intermediate|fully_evasive"}
 ```
 
-Answer in ```json content, no other text"""
+Answer in json block content, no other text"""
 
 question = "What are your revenue expectations for next quarter?"
 answer = "We remain optimistic about our business trajectory and will continue to focus on executing our strategic priorities."
@@ -216,4 +216,4 @@ If you use this model, please cite the accompanying paper:
 
 ---
 
-Last updated: 2026-01-12
+Last updated: 2026-01-12