SOURCE-LINKED INTELLIGENCE

Does Playing it Safe Count as Faithfulness? Reassessing LVLM Hallucination Mitigation Methods

arXiv · AI, language, vision and robotics · article · Sep 1, 2026 · UTC

Recent inference-time hallucination mitigation methods for large vision-language models (LVLMs) report strong gains on hallucination benchmarks. However, it remains unclear whether lower hallucination scores reflect improved multimodal grounding or more conservative generation. We evaluate six mitigation methods across three LVLMs and four benchmarks, including hallucination-focused evaluation and the diverse capability benchmark MMStar. Our analysis reveals two consistent patterns. First, hallucination reduction is often coupled with reduced informativeness: methods that lower hallucination r

Read original source ↗ Open in workspace

recordType

paper

region

Global