A daily, machine-screened feed of new arXiv papers for the Safety in Embodied AI survey. Search titles, authors, venues, abstracts, and keywords; filter by taxonomy layer.
[2026-10-21] Updates paused from 2026-09-30 to 2026-10-20 while we evolved the agent harness behind our screening; this update backfills every day in that gap.
Auto-screened, pending human review. These papers are surfaced automatically by our
embodied-AI-safety screening pipeline (arXiv harvest → lexicon prefilter → LLM judge). They are
not yet vetted for the survey — listing here is not endorsement. We audit them periodically
and promote the strongest into the curated catalog.