The Evidence-Grounded, Social-Weighted Persona Panel is proposed, a three-stage GenUI evaluation method in which a panel of psychologically diverse, evidence-grounded personas independently rates a screenshot, exchanges opinions under a trait-derived, semantically-gated bounded-confidence mechanism, and is aggregated v...
Zheng Wu, Yibo Luo, Pu Zhang et al.· arXiv.org· 0 citations
MineExplorer benchmark for evaluating open-world exploration capabilities of MLLM agents in Minecraft is introduced, which first filter atomic tasks whose solutions rely heavily on Minecraft-specific knowledge to better reflect general open-world reasoning and compose atomic tasks into implicit multi-hop tasks.
Results validate that a lightweight inference-time intervention can substantially boost GUI Agent's competence to identify inappropriate execution scenarios and refrain from unnecessary actions.
Zhao-Yuan Huang, Tian-Jie Ju, Pengzhou Cheng et al.· 1 citation
This work proposes a causal framework based on activation steering to actively probe and manipulate internal visual representations, revealing a divergence in concept encoding: entity knowledge is distinctively localized, whereas abstract concepts are globally distributed across the network.
This paper proposes UrbanGround, the first sandbox to make this question testable in a physically constrained replica of Hong Kong built from territory-wide 3D geospatial data, and hopes it will support broader study of how far current MLLM agents can explore reliably in complex, open-ended urban environments.
Tianjie Ju, Zheng Wu, Yue-Qing Sun et al.· 2 citations
The findings relocate the bottleneck of commonsense reasoning failures from model competence to elicitation, and release SaliTrap as a testbed for this blind spot, to show that lightweight, inference-time prompting alone substantially closes the gap without any retraining.
