Learning to Fold: prizewinning solution at LeHome Challenge 2026 (1st place online, 2nd offline)

  • 类型:arxiv
  • 标识:2606.27163
  • 链接:https://arxiv.org/abs/2606.27163
  • 主分类:multimodal
  • 形态:method
  • 被引:0
  • 被引来源:Semantic Scholar + OpenAlex
  • S2被引:0
  • OpenAlex被引:0
  • 影响力被引:0
  • TLDR:The work improves a vision-language-action (VLA) policy with a reinforcement-learning loop that predicts success, progress, and a few task-relevant future quantities and drives advantage estimation, live failure detection, and candidate selection in the LeHome Challenge 2026.
  • OpenAlex ID:W7166008352
  • OpenAlex DOI:10.48550/arxiv.2606.27163
  • DOI:10.48550/arxiv.2606.27163
  • DOI来源:OpenAlex
  • 开放获取:green
  • 开放获取链接:https://doi.org/10.48550/arxiv.2606.27163
  • OpenAlex更新:2026-07-19
  • 待LLM分类:否
  • 标题中文:Learning to Fold:LeHome Challenge 2026 获奖方案(线上第 1,线下第 2)
  • TLDR中文:本工作通过强化学习循环改进视觉-语言-动作(VLA)策略,该循环预测成功、进展及若干任务相关的未来量,并驱动优势估计、实时失败检测与候选选择,在 LeHome Challenge 2026 中取得佳绩。
  • 来源文件
  • /inbox/tom/_candidates/2026-06-29-agent-rag-longcontext-candidates.json
  • [S2 enrich]
  • [OpenAlex backfill]