DAIL: Beyond Task Ambiguity for Language-Conditioned Reinforcement Learning
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911227530182656 |
|---|---|
| author | Xie, Runpeng Wang, Quanwei Hu, Hao Zhou, Zherui Mu, Ni Li, Xiyun Yang, Yiqin Xu, Shuang Zhao, Qianchuan XU, Bo |
| author_facet | Xie, Runpeng Wang, Quanwei Hu, Hao Zhou, Zherui Mu, Ni Li, Xiyun Yang, Yiqin Xu, Shuang Zhao, Qianchuan XU, Bo |
| contents | Comprehending natural language and following human instructions are critical capabilities for intelligent agents. However, the flexibility of linguistic instructions induces substantial ambiguity across language-conditioned tasks, severely degrading algorithmic performance. To address these limitations, we present a novel method named DAIL (Distributional Aligned Learning), featuring two key components: distributional policy and semantic alignment. Specifically, we provide theoretical results that the value distribution estimation mechanism enhances task differentiability. Meanwhile, the semantic alignment module captures the correspondence between trajectories and linguistic instructions. Extensive experimental results on both structured and visual observation benchmarks demonstrate that DAIL effectively resolves instruction ambiguities, achieving superior performance to baseline methods. Our implementation is available at https://github.com/RunpengXie/Distributional-Aligned-Learning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_19562 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DAIL: Beyond Task Ambiguity for Language-Conditioned Reinforcement Learning Xie, Runpeng Wang, Quanwei Hu, Hao Zhou, Zherui Mu, Ni Li, Xiyun Yang, Yiqin Xu, Shuang Zhao, Qianchuan XU, Bo Artificial Intelligence Comprehending natural language and following human instructions are critical capabilities for intelligent agents. However, the flexibility of linguistic instructions induces substantial ambiguity across language-conditioned tasks, severely degrading algorithmic performance. To address these limitations, we present a novel method named DAIL (Distributional Aligned Learning), featuring two key components: distributional policy and semantic alignment. Specifically, we provide theoretical results that the value distribution estimation mechanism enhances task differentiability. Meanwhile, the semantic alignment module captures the correspondence between trajectories and linguistic instructions. Extensive experimental results on both structured and visual observation benchmarks demonstrate that DAIL effectively resolves instruction ambiguities, achieving superior performance to baseline methods. Our implementation is available at https://github.com/RunpengXie/Distributional-Aligned-Learning. |
| title | DAIL: Beyond Task Ambiguity for Language-Conditioned Reinforcement Learning |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2510.19562 |