Research Papers 论文研究 5h ago Updated 1h ago 更新于 1小时前 46

First Make It Playable, Then Make It Good: Staged Interaction Learning for Small Dialogue-Game Agents 先让它可玩,再让它好玩:小型对话游戏代理的分阶段交互学习

Qwen-GuidePlay-2B is a 2B-parameter dialogue-game agent fine-tuned from Qwen3.5-2B using a staged three-step SFT pipeline The approach achieves 57.12 clemscore and 42.68 statscore on Playpen validation, with a +36 clemscore delta over the base model in official challenge results (second-highest among all submitted systems) The three-stage training consists of: successful-trajectory SFT, weighted turn-level SFT, and teacher-guided SFT where the teacher only fixes formatting and evaluates examples 提出Qwen-GuidePlay-2B,一个仅2B参数的对话游戏交互语言模型,基于Qwen3.5-2B分三阶段微调 采用"先可玩后优化"策略:成功轨迹SFT→加权回合级SFT→教师引导SFT,教师模型仅用于格式修复和评估 在Playpen验证集上获得57.12 clemscore和42.68 statscore,challenge结果中clemscore delta排名第二(约+36) 发现小模型性能提升依赖精心数据策展而非激进算法,重程序方法(replay-repair、hard-example mining)无效 模型与代码已开源,支持复现

62
Hot 热度
72
Quality 质量
65
Impact 影响力

Analysis 深度分析

TL;DR

  • Qwen-GuidePlay-2B is a 2B-parameter dialogue-game agent fine-tuned from Qwen3.5-2B using a staged three-step SFT pipeline
  • The approach achieves 57.12 clemscore and 42.68 statscore on Playpen validation, with a +36 clemscore delta over the base model in official challenge results (second-highest among all submitted systems)
  • The three-stage training consists of: successful-trajectory SFT, weighted turn-level SFT, and teacher-guided SFT where the teacher only fixes formatting and evaluates examples without generating new gold actions
  • Procedurally heavy methods like replay-repair and hard-example mining provided no benefit, suggesting careful data curation is more effective than aggressive algorithmic changes for small models
  • Both the model and code are released for reproducibility

Why It Matters

This work demonstrates that small language models (2B parameters) can achieve competitive performance in complex dialogue-game settings through a carefully designed staged fine-tuning strategy, challenging the assumption that only large models can handle interactive tasks effectively. The findings offer a practical blueprint for practitioners looking to build efficient dialogue agents without relying on massive compute or procedurally expensive training techniques.

Technical Details

  • Base model: Qwen3.5-2B, fine-tuned via three sequential SFT stages on the Playpen dialogue-game dataset
  • Stage 1: Supervised fine-tuning on only successful game trajectories to establish baseline playability
  • Stage 2: Weighted turn-level SFT to improve per-turn decision-making quality
  • Stage 3: Teacher-guided SFT using a larger model solely for formatting correction and example evaluation — the teacher does not generate new gold-standard actions, avoiding distribution shift
  • Benchmarks: Playpen validation set with clemscore (57.12) and statscore (42.68); official challenge results showed second-highest clemscore delta (~+36 over base)
  • Negative findings: Replay-repair and hard-example mining approaches did not improve performance, indicating that data curation quality outweighs procedural complexity for small models

Industry Insight

  • Small models can be highly competitive in interactive dialogue tasks when trained with staged, curriculum-style fine-tuning — organizations should consider investing in data curation pipelines rather than scaling model size alone
  • The teacher-only-formatting approach avoids the risk of policy drift that can occur when larger models generate training labels, offering a safer distillation paradigm for production systems
  • The failure of replay-repair and hard-example mining suggests that for resource-constrained settings, simpler and more deliberate training strategies may outperform complex algorithmic interventions, redirecting engineering effort toward data quality over training sophistication

TL;DR

  • 提出Qwen-GuidePlay-2B,一个仅2B参数的对话游戏交互语言模型,基于Qwen3.5-2B分三阶段微调
  • 采用"先可玩后优化"策略:成功轨迹SFT→加权回合级SFT→教师引导SFT,教师模型仅用于格式修复和评估
  • 在Playpen验证集上获得57.12 clemscore和42.68 statscore,challenge结果中clemscore delta排名第二(约+36)
  • 发现小模型性能提升依赖精心数据策展而非激进算法,重程序方法(replay-repair、hard-example mining)无效
  • 模型与代码已开源,支持复现

为什么值得看

本文展示了小参数模型在对话游戏任务上的高效微调策略,为资源受限场景下的Agent开发提供了可复现的范式。其"分阶段学习"思路对构建可玩性强且决策质量高的对话系统具有直接参考价值。

技术解析

  • 模型架构:基于Qwen3.5-2B,参数量仅2B,专为对话游戏交互设计
  • 三阶段微调策略:①仅在Playpen成功轨迹上进行SFT,确保基础可玩性;②加权回合级SFT,优化单回合决策质量;③教师引导SFT,大模型仅用于格式修正和示例评估,不生成新gold actions
  • 评估指标:clemscore(57.12)衡量对话流畅性,statscore(42.68)衡量任务完成度
  • 负结果发现:replay-repair和hard-example mining等重程序方法未带来提升,表明小模型瓶颈在于数据质量而非算法复杂度
  • 开源情况:模型权重与训练代码均已公开

行业启示

  • 小模型在特定任务上可通过精细化数据策展实现显著性能提升,无需盲目追求大参数或复杂训练技巧
  • "先保证可玩性再优化决策"的分阶段策略可作为对话Agent开发的通用范式,降低训练风险
  • 教师模型应谨慎使用:仅作为格式/评估工具而非内容生成器,可避免引入分布偏移

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

LLM 大模型 Fine-tuning 微调 Agent Agent Conversational AI 对话系统 Research 科学研究