Backend
NONSENSE: Non-Semantic Strings Expose Semanticity Gaps in Dense Retrieval
arXiv cs.IR0 views
arXiv:2611.22353v1 Announce Type: new Abstract: Dense retrieval is central to modern search, question answering, and retrieval-augmented generation, where high similarity is often treated as evidence of semantic relevance. We examine how far this assumption can be stressed. NONSENSE is a black-box sparse-edit probe that constructs visibly non-semantic strings while forbidding query-term copying, optimizing only the dense similarity between a query and the string. Across five BEIR-style datasets and three popular single-vector retrievers, these strings frequently outscore the retriever's own best judged query-relevant passage (QRP), and often rank first in the full corpus. The result indicates a semanticity gap in the evaluated dense retrievers: nearest-neighbor status can be separated from passage-like meaning. These findings highlight the importance of evaluating dense retrievers not only by whether they retrieve relevant documents, but also by how reliably their similarity scores reflect semantic evidence.
Source:https://arxiv.org/abs/2611.22353PreviousTeaching Language Models to Narrow Boolean Patent Searches from Search FeedbackNextDeal With It: On Better Schedules for Listwise Reranking
Related posts

Java hit by major security flaw — here's what we know about Java4Shell so far
0TechRadar
Quiz: Nuitka: Compile Python Into a Standalone Executable
0Real Python
Why Microsoft's Outlook Outage Lasted More Than 27 Hours
0HackerNoon

I Watched AI Eat the Easy Half of an Industry. The Other Half Got More Expensive.
0HackerNoon
AI Agents Could Drive the Next Open Source Software Revolution
0HackerNoon
Prompt Theory: Critical Analysis
0dev.to