doc: mentioned new models used for evaluation

README.md

@@ -218,7 +218,7 @@ Success requires the model to utilize the hidden entity as a semantic bridge:
 
 ### Key Features
 *   **Adversarially Hardened:** Unlike standard MCQs, ShadowBench (v3) is filtered to prevent "shortcut learning" via gendered pronouns, chronological era-matching, or category-leaks.
-*   **Scale Robust:** Evaluated on models ranging from 8B parameters (Llama-3, Qwen3) to frontier scales (GPT-5.4).
+*   **Scale Robust:** Evaluated on models ranging from 8B parameters (Llama-3, Qwen3) to frontier scales (GPT-5.4-mini, GPT-5.4, and Claude-Sonnet-4.6).
 *   **Multi-Domain:** Covers Technology, Sports (Tennis), and Entertainment (Actors).
 *   **Stratified:** Includes "Upper Tier" (Head) and "Lower Tier" (Tail) entities based on Wikipedia popularity metrics to evaluate "Popularity Bias."