FINER: MLLMs Hallucinate under Fine-grained Negative Queries
This addresses a critical challenge in MLLMs for applications requiring precise image understanding, though it is incremental as it builds on existing methods like DPO.
The paper tackles the problem of hallucinations in multimodal large language models (MLLMs) under fine-grained negative queries, introducing the FINER benchmarks to analyze this issue and proposing FINER-Tuning, which reduces hallucinations by up to 24.2% on their benchmarks while improving performance on existing suites.
Multimodal large language models (MLLMs) struggle with hallucinations, particularly with fine-grained queries, a challenge underrepresented by existing benchmarks that focus on coarse image-related questions. We introduce FIne-grained NEgative queRies (FINER), alongside two benchmarks: FINER-CompreCap and FINER-DOCCI. Using FINER, we analyze hallucinations across four settings: multi-object, multi-attribute, multi-relation, and ``what'' questions. Our benchmarks reveal that MLLMs hallucinate when fine-grained mismatches co-occur with genuinely present elements in the image. To address this, we propose FINER-Tuning, leveraging Direct Preference Optimization (DPO) on FINER-inspired data. Finetuning four frontier MLLMs with FINER-Tuning yields up to 24.2\% gains (InternVL3.5-14B) on hallucinations from our benchmarks, while simultaneously improving performance on eight existing hallucination suites, and enhancing general multimodal capabilities across six benchmarks. Code, benchmark, and models are available at \href{https://explainableml.github.io/finer-project/}{https://explainableml.github.io/finer-project/}.