Large Language Models (LLMs) have achieved SOTA results in many natural language pro- cessing tasks. However, they are primarily trained for static text generation and often strug- gle in interactive, goal-oriented environments that require multi-turn reasoning, state track- ing, and need to give structured output formats. This paper presents our team’s submission to the LM Playschool Workshop Shared Task at EMNLP 2026. The LM Playschool Workshop Shared Task at EMNLP 2026 addresses this challenge by evaluating language models as in- teractive agents in structured dialogue games. We hereby explored different fine-tuning strate- gies for Qwen 4B and Qwen 7B models to improve their interactive capabilities and eval- uated them. We perform a comparison of Su- pervised Fine-Tuning (SFT) and Direct Pref- erence Optimization (DPO) using multi-turn training data. Our study analyzes the effective- ness of each optimization strategy for learning robust dialogue behaviors while maintaining models stability towards interactive games and also not to forget its language linguistics by different evaluating metrics. Experimental re- sults show that although supervised fine-tuning provides consistent interactive performance, di- rectly optimizing the foundation model using DPO reduces optimization errors and leads to improved agent behavior. Our best-performing model achieved a clemscore of 38.23 on the official LM Playschool leaderboard, evaluated using the Playpen workbench.