Enregistré dans:
Détails bibliographiques
Auteurs principaux: Huang, Yitian, Lei, Yuxuan, Lian, Jianxun, Liao, Hao
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2511.20200
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866908675274178560
author Huang, Yitian
Lei, Yuxuan
Lian, Jianxun
Liao, Hao
author_facet Huang, Yitian
Lei, Yuxuan
Lian, Jianxun
Liao, Hao
contents This report presents the solution and results of our team MSRA\_SC in the Commonsense Persona-Grounded Dialogue Challenge (CPDC 2025). We propose a simple yet effective framework that unifies improvements across both GPU Track and API Track. Our method centers on two key components. First, Context Engineering applies dynamic tool pruning and persona clipping for input compression, combined with post-processing techniques such as parameter normalization and function merging. Together with manually refined prompts, this design improves tool call stability, execution reliability, and role-playing guidance. Second, in the GPU Track, we further adopt GRPO training, replacing supervised fine-tuning with reinforcement learning directly optimized by reward signals. This mitigates small-sample overfitting and significantly enhances task-oriented dialogue performance. In the final evaluation, our team ranks 1st in Task 2 API, 2nd in Task 1 API, and 3rd in both Task 3 API and GPU track, demonstrating the effectiveness of our approach. Our code is publicly available at https://gitlab.aicrowd.com/nikoo_yu/cpdc-2025-winning-solution
format Preprint
id arxiv_https___arxiv_org_abs_2511_20200
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Interactive AI NPCs Powered by LLMs: Technical Report for the CPDC Challenge 2025
Huang, Yitian
Lei, Yuxuan
Lian, Jianxun
Liao, Hao
Artificial Intelligence
This report presents the solution and results of our team MSRA\_SC in the Commonsense Persona-Grounded Dialogue Challenge (CPDC 2025). We propose a simple yet effective framework that unifies improvements across both GPU Track and API Track. Our method centers on two key components. First, Context Engineering applies dynamic tool pruning and persona clipping for input compression, combined with post-processing techniques such as parameter normalization and function merging. Together with manually refined prompts, this design improves tool call stability, execution reliability, and role-playing guidance. Second, in the GPU Track, we further adopt GRPO training, replacing supervised fine-tuning with reinforcement learning directly optimized by reward signals. This mitigates small-sample overfitting and significantly enhances task-oriented dialogue performance. In the final evaluation, our team ranks 1st in Task 2 API, 2nd in Task 1 API, and 3rd in both Task 3 API and GPU track, demonstrating the effectiveness of our approach. Our code is publicly available at https://gitlab.aicrowd.com/nikoo_yu/cpdc-2025-winning-solution
title Interactive AI NPCs Powered by LLMs: Technical Report for the CPDC Challenge 2025
topic Artificial Intelligence
url https://arxiv.org/abs/2511.20200