Middle-layer contextual embeddings, not logit-lens readings, improve hallucination detection in VLMs and enable bounding-box grounding for visual question answering.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Beyond Logit Lens: Contextual Embeddings for Robust Hallucination Detection & Grounding in VLMs
Middle-layer contextual embeddings, not logit-lens readings, improve hallucination detection in VLMs and enable bounding-box grounding for visual question answering.