Additional Paper for the Hallucination Benchmarks section

[AtsuMiyai]

Hello. Thanks for your comprehensive survey on MLLM's hallucination, which is really beneficial in this field.

We published a paper that tackled a novel task called Unsolvable Problem Detection (UPD) for MLLM. https://arxiv.org/abs/2403.20331

UPD examines the MLLM’s ability to withhold answers when faced with unsolvable problems. We also propose MM-UPD Bench and benchmark recent MLLMs with MM-UPD Bench.

We consider the definition and practical evaluation of the specific UPD problem to provide a novel and important contribution to existing hallucination benchmarks.

Could you add our work to the Hallucination Benchmarks section?

• Unsolvable Problem Detection Unsolvable Problem Detection: Evaluating Trustworthiness of Vision Language Models (Mar. 29, 2024)

[JosephPai]

Hi, the paper has just been updated to the benchmark section. Thanks for your interests!