Beyond Conversations: Spatially-Anchored Previews for Intent Disambiguation in LLM-Assisted Geometry Editing in Virtual Reality
3.40T1 sourcearXiv cs.HC
Source record
Published by arXiv cs.HC (T1 source). The original is at https://arxiv.org/abs/2607.23201.
Pipeline notes
The summary and note below are generated by the signal pipeline — they are Beyond Desk’s reading, not quotations from the source.
SummaryWithin-subjects study (n=24) evaluating spatially-anchored graphical previews combined with clarification questions for disambiguating user intent in LLM-assisted parameter-driven geometry editing in VR. The hybrid approach reduced conversation rounds, improved interaction stability, and enhanced user experience versus no disambiguation.
Why it mattersQuantified disambiguation design choices for LLM editing tools in immersive environments, with measured reductions in dialogue overhead and gains in interaction stability. Niche but useful as a reference for VR/AR agent-tool builders.
Cited by
No citations on record.