Source record

Published by arXiv cs.HC (T1 source). The original is at https://arxiv.org/abs/2609.28812.

Pipeline notes

The summary and note below are generated by the signal pipeline — they are Beyond Desk’s reading, not quotations from the source.

Within-subjects study (N=24) comparing three designs for presenting hallucination-awareness cues—uncertainty and provenance—in VR embodied conversational agents against a no-cue baseline. Embodied cues increased trust and immersion; text cues improved interpretability; icon cues balanced interpretability with minimal immersion disruption.

Empirical, design-comparative evidence on how to signal LLM uncertainty in immersive agent interfaces, with concrete tradeoffs in trust, interpretability, and immersion.

Cited by

No citations on record.