Metraipolis › archive › dc_e2f04f874914
Do Better Goal Representations Improve Goal-Conditioned Reinforcement Learning? — arXiv:2609.39901
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.39901
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T12:45:52.664Z
- SHA-256 of the archived text
07270defe990f1b4cb16d0f1e3363a963e3419500d1f47d80bc939db159d4d18- Stable URL of this page
https://metraipolis.com/documento/dc_e2f04f874914
How to cite
Do Better Goal Representations Improve Goal-Conditioned Reinforcement Learning? — arXiv:2609.39901 — arXiv (en), https://arxiv.org/abs/2609.39901 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e2f04f874914 (Metraipolis archive dc_e2f04f874914, sha256 07270defe990f1b4cb16d0f1e3363a963e3419500d1f47d80bc939db159d4d18)
BibTeX
@misc{dce2f04f874914,
title = {Do Better Goal Representations Improve Goal-Conditioned Reinforcement Learning? — arXiv:2609.39901},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.39901},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e2f04f874914; sha256 07270defe990f1b4cb16d0f1e3363a963e3419500d1f47d80bc939db159d4d18},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.39901. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Goal-conditioned reinforcement learning (GCRL) relies heavily on how target goals are represented to the policy. While recent methods encode goals via temporal distance, occupancy, or controllability, it remains unclear how much downstream performance actually depends on representation quality. We study this in offline GCRL by constructing an exact temporal-distance goal representation in deterministic mazes. We then systematically corrupt its geometric quality while keeping the downstream learner fixed. Across OGBench navigation tasks and two algorithms, large changes in goal-representation quality produce almost no change in performance. However, applying the same interventions to the agent's current state more than doubles success, revealing the state pathway as the true bottleneck. Building on this insight, we show that simple random Fourier positional encodings substantially improve performance on the hardest navigation tasks without map information or objective modifications. Overall, our findings suggest that in state-based offline navigation, improving how the agent's current state is represented matters far more than refining the goal representation. Code will be released soon. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-10-01 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.39901 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-07T12:14:14.168Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-10-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).