Update README.md

README.md

@@ -20,9 +20,9 @@ model_type: llama
 
 ##  Overview
 
-This model is a customized instruction-following LLM optimized for **long-context dialogue generation**, especially in **emotionally intense**, **personified**, or **roleplay-based applications**. It supports unfiltered text outputs, character-level memory continuity, and emotion-driven response modulation—ideal for developing **private companion bots**, **NSFW story agents**, and **synthetic relationship simulations**.
+This model is a customized instruction-following LLM optimized for long-context dialogue generation, especially in emotionally intense, personified**, or roleplay-based applications. It supports unfiltered text outputs, character-level memory continuity, and emotion-driven response modulation—ideal for developing private companion bots, NSFW story agents, and synthetic relationship simulations.
 
-The design draws architectural inspiration from open-source efforts in **conversational memory**, **persona conditioning**, and **preference-aligned decoding strategies**, as well as from user interaction behaviors observed in existing virtual AI companion platforms.
+The design draws architectural inspiration from open-source efforts in conversational memory, persona conditioning, and preference-aligned decoding strategies, as well as from user interaction behaviors observed in existing virtual AI companion platforms.
 
 
 ## Technical Highlights