REVIEW 6 cited by
Investigating and Mitigating the Multimodal Hallucination Snowballing in Large Vision-Language Models
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
abstract
Though advanced in understanding visual information with human languages, Large Vision-Language Models (LVLMs) still suffer from multimodal hallucinations. A natural concern is that during multimodal interaction, the generated hallucinations could influence the LVLMs' subsequent generation. Thus, we raise a question: When presented with a query relevant to the previously generated hallucination, will LVLMs be misled and respond incorrectly, even though the ground visual information exists? To answer this, we propose a framework called MMHalSnowball to evaluate LVLMs' behaviors when encountering generated hallucinations, where LVLMs are required to answer specific visual questions within a curated hallucinatory conversation. Crucially, our experiment shows that the performance of open-source LVLMs drops by at least $31\%$, indicating that LVLMs are prone to accept the generated hallucinations and make false claims that they would not have supported without distractions. We term this phenomenon Multimodal Hallucination Snowballing. To mitigate this, we further propose a training-free method called Residual Visual Decoding, where we revise the output distribution of LVLMs with the one derived from the residual visual input, providing models with direct access to the visual information. Experiments show that our method can mitigate more than $24\%$ of the snowballed multimodal hallucination while maintaining capabilities.
Forward citations
Cited by 6 Pith papers
-
CLAIM: Mitigating Multilingual Object Hallucination in Large Vision-Language Models with Cross-Lingual Attention Intervention
An inference-time attention-shift intervention aligns non-English queries' cross-modal attention with English, cutting multilingual object hallucination in LVLMs on POPE and MME.
-
One for All: Update Parameterized Knowledge Across Multiple Models
One fine-tuned small model plus an ensemble step can update a fact across multiple large language models with a single edit, outperforming separate per-model editing.
-
OmniDPO: A Preference Optimization Framework to Address Omni-Modal Hallucination
OmniDPO extends direct preference optimization with audio-video alignment and modality-degradation preference pairs to reduce omni-modal hallucination.
-
CAI: Caption-Sensitive Attention Intervention for Mitigating Object Hallucination in Large Vision-Language Models
CAI reduces object hallucination in LVLMs by injecting caption-query attention patterns into selected attention heads at inference time.
-
Seeing Far and Clearly: Mitigating Hallucinations in MLLMs with Attention Causal Decoding
FarSight adds upper-triangular negative biases to the causal mask to absorb outlier-token attention, reducing hallucinations in MLLMs without training.
-
Empowering Multimodal LLMs with External Tools: A Comprehensive Survey
A survey paper maps how external tools are used to augment multimodal large language models across data, tasks, evaluation, and future directions.
Discussion (0). Continue with ORCID to comment.