Open-World Multi-Task Control Through Goal-Aware Representation Learning and Adaptive Horizon Prediction
Open accessAbstract
We study the problem of learning goal-conditioned policies in Minecraft, a popular, widely accessible yet challenging open-ended environment for developing human-level multi-task agents. We first identify two main challenges of learning such policies: 1) the indistinguishability of tasks from the state distribution, due to the vast scene diversity, and 2) the non-stationary nature of environment dynamics caused by partial observability. To tackle the first challenge, we propose Goal-Sensitive Backbone (GSB) for the policy to encourage the emergence of goal-relevant visual state representations. To tackle the second challenge, the policy is further fueled by an adaptive horizon prediction module that helps alleviate the learning uncertainty brought by the non-stationary dynamics. Experiments on 20 Minecraft tasks show that our method significantly outperforms the best baseline so far; in many of them, we double the performance. Our ablation and exploratory studies then explain how our approach beat the counterparts and also unveil the surprising bonus of zero-shot generalization to new scenes (biomes). We hope our agent could help shed some light on learning goal-conditioned, multi-task agents in challenging, open-ended environments like Minecraft. The code is released at https://github.com/CraftJarvis/MC-Controller.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
It would not be an apple-to-apple comparison to directly compare with prior methods that take Minecraft screen pixels as input and output low-level controls, because VOYAGER relies on the high-level Mineflayer API to control the agent.
“Note that we do not directly compare with prior methods that take Minecraft screen pixels as input and output low-level controls [53-55]. It would not be an apple-to-apple comparison, because we rely on the high-level Mineflayer [52] API to control the agent.”
From Voyager: An Open-Ended Embodied Agent with Large Language Models (2023) · cites this paper as [54]
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474 · see citing works
Film: Visual reasoning with a general conditioning layer
Perez, E., Strub, F., De Vries, H., Dumoulin, V., & Courville, A. (2018) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v32i1.11671 · Cited by 2,490 · see citing works
BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based Masked Language-models
Ben Zaken, E., Goldberg, Y., & Ravfogel, S. (2022) · Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers) · doi:10.18653/v1/2022.acl-short.1 · Cited by 765 · see citing works
Do As I Can, Not As I Say: Grounding Language in Robotic Affordances
Ahn, M., Brohan, A., Brown, N., Chebotar, Y., Cortes, O., David, B., Finn, C., Fu, C., Gopalakrishnan, K., Hausman, K., et al. (2022) · arXiv · doi:10.48550/arxiv.2204.01691 · Cited by 624 · see citing works
MineRL: A Large-Scale Dataset of Minecraft Demonstrations
Guss, W. H., Houghton, B., Topin, N., Wang, P., Codel, C., Veloso, M., & Salakhutdinov, R. (2019) · Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence · doi:10.24963/ijcai.2019/339 · Cited by 94 · see citing works
The MineRL 2019 Competition on Sample Efficient Reinforcement Learning using Human Priors
Guss, W. H., Codel, C., Hofmann, K., Houghton, B., Kuno, N., Milani, S., Mohanty, S., Liebana, D. P., Salakhutdinov, R., Topin, N., et al. (2019) · arXiv · doi:10.48550/arxiv.1904.10079 · Cited by 46 · see citing works
MineRL Diamond 2021 Competition: Overview, Results, and Lessons Learned
Kanervisto, A., Milani, S., Ramanauskas, K., Topin, N., Lin, Z., Li, J., Shi, J., Ye, D., Fu, Q., Yang, W., et al. (2022) · arXiv · doi:10.48550/arxiv.2202.10583 · Cited by 6 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Voyager: An Open-Ended Embodied Agent with Large Language Models
Wang, G., Xie, Y., Jiang, Y., Mandlekar, A., Xiao, C., Zhu, Y., Fan, L., & Anandkumar, A. (2023) · arXiv · doi:10.48550/arxiv.2305.16291 · Cited by 222 · see citing works
“[55] Shaofei Cai, Zihao Wang, Xiaojian Ma, Anji Liu, and Yitao Liang.”
Background<b>JARVIS</b> -1: Open-World Multi-Task Agents With Memory-Augmented Multimodal Language Models
Wang, Z., Cai, S., Liu, A., Jin, Y., Hou, J., Zhang, B., Lin, H., He, Z., Zheng, Z., Yang, Y., et al. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2024.3511593 · Cited by 36 · see citing works
“Compared to canonical scenarios with relatively small scale, simple dynamics, and limited tasks, open-world environments impose substantial challenges to building agents that can accomplish a diverse set of tasks [14, 18, 17, 23, 9, 49, 10].”
Self-citation
- Authors
- Cai, S., Wang, Z., Ma, X., Liu, A., & Liang, Y.
- Affiliations
- Peking University (CN)
- University of California, Los Angeles (US)
- Venue
- 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
- Publisher
- IEEE
- Published
- 2023-06
- Pages
- 13734-13744
- ISSN
- 1063-6919 (ISSN-L)
- OpenAlex
- W4386065365
- Cited by
- 12
- References
- 7
- Citations by year
- 2026: 1 · 2025: 4 · 2024: 6 · 2023: 1
- FWCI
- 2.13
- Type
- conference paper
- Language
- en
- Open access
- green
- arxiv.org
- Locations
- IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings (publishedVersion)
- arXiv (Cornell University) (submittedVersion, cc-by)
- arXiv (Cornell University) (cc-by)
- Funders
- National Key R&D Program of China (2022ZD0160301)
- NSF (IIS-1943641,IIS-1956441,CCF-1837129)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence in Games
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Keywords
- goal-conditioned policies, multi-task learning, open-ended environments, Minecraft, representation learning, non-stationary dynamics, partial observability, zero-shot generalization
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
