{"slug": "generative-embodied-multiple-behavior-control-systems-for-human-like-agents", "title": "Generative Embodied Multiple Behavior Control Systems for Human-like Agents", "summary": "A new arXiv paper (2609.22691v1) proposes a human behavior control framework that jointly models goal-directed and habitual behaviors for embodied human-like agents, with a Habitual Controller retrieving cue-triggered behaviors from personalized habit memory and a Goal-directed Controller using a context-aware world model to predict action consequences and estimate their values. An Arbiter dynamically balances the two systems according to individual differences and momentary internal states, and a keyframe-guided 3D motion generation module reconstructs diverse human-level behavior instructions in 3D environments. The authors report that human-likeness performance was significantly improved in evaluations, human studies, and ablation studies, indicating benefits from leveraging habitual behavior and multiple behavior control system coordination for believable embodied human-like agents.", "body_md": "arXiv:2609.22691v1 Announce Type: new \nAbstract: An enduring and richly elaborated dichotomy in cognitive neuroscience is that of human behavior control mechanisms, divided into habitual versus goal-directed. While existing human-like agent frameworks primarily focus on modeling goal- directed behavior, habitual behavior has been largely overlooked, though it plays a crucial role in human daily life. In this paper, we address this gap by studying multiple behavior control systems that jointly model goal-directed and habitual behaviors. We propose a human behavior control mechanism-inspired framework which the Habitual Controller retrieves cue-triggered behaviors from personal- ized habit memory, while the Goal-directed Controller employs a context-aware world model to predict action consequences and estimate their values. The Arbiter dynamically balances the influence of both systems according to individual differ- ences and momentary internal states. To reconstruct diverse human-level behavior instructions in 3D environments, we further develop a keyframe-guided 3D mo- tion generation module. Through extensive evaluation methods, human studies, and ablations studies, experimental results demonstrate that human-likeness per- formance is significantly improved by our approach. The efficacy of our approach indicates the benefits of leveraging habitual behavior and multiple behavior con- trol system coordination for believable embodied human-like agents.", "url": "https://wpnews.pro/news/generative-embodied-multiple-behavior-control-systems-for-human-like-agents", "canonical_source": "https://arxiv.org/abs/2609.22691", "published_at": "2026-09-23 04:00:00+00:00", "updated_at": "2026-09-23 04:26:56.273724+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-agents", "ai-research", "machine-learning"], "entities": ["arXiv", "Habitual Controller", "Goal-directed Controller", "Arbiter"], "alternates": {"html": "https://wpnews.pro/news/generative-embodied-multiple-behavior-control-systems-for-human-like-agents", "markdown": "https://wpnews.pro/news/generative-embodied-multiple-behavior-control-systems-for-human-like-agents.md", "text": "https://wpnews.pro/news/generative-embodied-multiple-behavior-control-systems-for-human-like-agents.txt", "jsonld": "https://wpnews.pro/news/generative-embodied-multiple-behavior-control-systems-for-human-like-agents.jsonld"}}