Autonomous Learning From Success and Failure: Goal-Conditioned Supervised Learning with Negative Feedback
ACM Transactions on Autonomous and Adaptive Systems · 20.11.2026
DOI: 10.1145/3864439 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Transactions on Autonomous and Adaptive Systems, online erschienen am 20.11.2026. Verfasst von Zeqiang Zhang und 4 weiteren Autor:innen. Die Arbeit nennt 80 Literaturangaben.
Abstract (Originalsprache)
Learning from reward functions or imitating human demonstrations requires human specification for every behaviour, a problem for long-lived self-adaptive systems whose goals and conditions cannot be fully anticipated at design time. Goal-Conditioned Supervised Learning (GCSL) offers a self-supervised alternative: agents imitate their own experience under hindsight-relabelled goals. However, learning only from self-generated data can reinforce an agent's initial biases, and relabelling focuses solely on successes, so the agent cannot learn from its mistakes. We propose Goal-Conditioned Supervised Learning with Negative Feedback (GCSL-NF), which evaluates each trajectory twice: positively with respect to relabelled goals, and correctively with respect to the original goal. The corrective target comes from a similarity function learned contrastively from trajectory-induced neighbourhood relations, so no reward function or geometric distance needs to be specified. Experiments show that GCSL-NF overcomes initial biases, benefits increasingly from negative feedback as learning progresses, and matches or surpasses leading GCSL- and HER-based methods on discrete and continuous goal-reaching tasks. Its learning signal remains informative under corrupted priors, runtime goal shifts, sensor and actuation noise, and constraints enforced only during training. By reducing reliance on prespecified rewards, GCSL-NF suits self-adaptive systems whose objectives are diverse, changing, or hard to engineer.
Zitieren
APA
Zhang, Z., Wurzberger, F., Schmid, G., Gottwald, S., & Braun, D. A. (2026). Autonomous Learning From Success and Failure: Goal-Conditioned Supervised Learning with Negative Feedback. ACM Transactions on Autonomous and Adaptive Systems. https://doi.org/10.1145/3864439
Vancouver
Zhang Z, Wurzberger F, Schmid G, Gottwald S, Braun DA. Autonomous Learning From Success and Failure: Goal-Conditioned Supervised Learning with Negative Feedback. ACM Transactions on Autonomous and Adaptive Systems. 2026. doi:10.1145/3864439
BibTeX
@article{Zhang2026,
author = {Zhang, Zeqiang and Wurzberger, Fabian and Schmid, Gerrit and Gottwald, Sebastian and Braun, Daniel A.},
title = {Autonomous Learning From Success and Failure: Goal-Conditioned Supervised Learning with Negative Feedback},
journal = {ACM Transactions on Autonomous and Adaptive Systems},
year = {2026},
doi = {10.1145/3864439}
}
EUVIMED