REVIEW 6 cited by
Modality-Fair Preference Optimization for Trustworthy MLLM Alignment
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
read the original abstract
Multimodal large language models (MLLMs) have achieved remarkable success across various tasks. However, separate training of visual and textual encoders often results in a misalignment of the modality. Such misalignment may lead models to generate content that is absent from the input image, a phenomenon referred to as hallucination. These inaccuracies severely undermine the trustworthiness of MLLMs in real-world applications. Despite attempts to optimize text preferences to mitigate this issue, our initial investigation indicates that the trustworthiness of MLLMs remains inadequate. Specifically, these models tend to provide preferred answers even when the input image is heavily distorted. Analysis of visual token attention also indicates that the model focuses primarily on the surrounding context rather than the key object referenced in the question. These findings highlight a misalignment between the modalities, where answers inadequately leverage input images. Motivated by our findings, we propose Modality-Fair Preference Optimization (MFPO), which comprises three components: the construction of a multimodal preference dataset in which dispreferred images differ from originals solely in key regions; an image reward loss function encouraging the model to generate answers better aligned with the input images; and an easy-to-hard iterative alignment strategy to stabilize joint modality training. Extensive experiments on three trustworthiness benchmarks demonstrate that MFPO significantly enhances the trustworthiness of MLLMs. In particular, it enables the 7B models to attain trustworthiness levels on par with, or even surpass, those of the 13B, 34B, and larger models.
Forward citations
Cited by 6 Pith papers
-
Mitigating Hallucinations in Multimodal LLMs via Object-aware Preference Optimization
CHAIR-DPO labels DPO preference pairs with CHAIR hallucinations computed from detector outputs, reducing object hallucinations in LLaVA models without proprietary judges.
-
V2T-CoT: From Vision to Text Chain-of-Thought for Medical Reasoning and Diagnosis
V2T-CoT combines visual region grounding with LLM-generated text rationale training to improve medical visual question answering accuracy and interpretability on four benchmarks.
-
Fast or Slow? Integrating Fast Intuition and Deliberate Thinking for Enhancing Visual Question Answering
FOCUS improves VQA accuracy by routing easy questions through fast zero-shot answering and hard questions through question-conditioned image segmentation before the final answer.
-
HSCR: Hierarchical Self-Contrastive Rewarding for Aligning Medical Vision Language Models
HSCR uses visual token dropout and logit contrast to construct self-generated dispreferred answers, then trains a medical VLM with explicit and implicit preference losses, improving zero-shot Rad-VQA, SLAKE, and PathV...
-
Zooming from Context to Cue: Hierarchical Preference Optimization for Multi-Image MLLMs
Context-to-Cue Direct Preference Optimization (CcDPO) reduces multi-image hallucinations in 7B multimodal LLMs by training on perturbed full-sequence captions and region-focused visual prompts, improving average multi...
-
Knowing or Guessing? Robust Medical Visual Question Answering via Joint Consistency and Contrastive Learning
RoMed and CCL: a 144k-question perturbation benchmark for medical VQA and a consistency-plus-contrastive training method that improves LLaVA-Med's accuracy and reduces answer variation.
Discussion (0). Continue with ORCID to comment.