List your company

Object Hallucination Detection on nocaps FOIL (Out-Domain)

89.1APEXPERT
69.02874.23979.4584.661May 20, 2026

Evaluation results

MethodLinks
EXPERTMethod Category=LLM-based2026.0589.190.898.7
BEiTScore-L/14Method Category=Encoder-based, Backbone=BEiT-L/142026.05878898.3
VQAScoreMethod Category=LLM-based, Backbone=CLIP-FlanT5-XXL2026.0586.887.794.6
FLEURMethod Category=LLM-based2026.0585.28796.2
BEiTScore-B/32Method Category=Encoder-based, Backbone=BEiT-B/322026.0584.585.197.2
LongCLIP-SMethod Category=Encoder-based, Backbone=ViT-L2026.0581.983.497.5
PAC-S++Method Category=Encoder-based, Backbone=ViT-L2026.0579.882.797.1
PAC-SMethod Category=Encoder-based, Backbone=ViT-B2026.0578.781.994.6
LongCLIP-SMethod Category=Encoder-based, Backbone=ViT-B2026.057779.695.8
PAC-S++Method Category=Encoder-based, Backbone=ViT-B2026.0576.980.793.7
PAC-SMethod Category=Encoder-based, Backbone=ViT-L2026.0576.279.993.4
BEiTScore-L/14 w/o Stage 1Method Category=Encoder-based, Backbone=BEiT-L/14, Training Stage=w/o Stage 12026.0576.278.595.5
CLIP-SMethod Category=Encoder-based, Backbone=ViT-L2026.0574.678.292.5
CLIP-SMethod Category=Encoder-based, Backbone=ViT-B2026.0573.878.689.5
BEiTScore-B/32 w/o Stage 1Method Category=Encoder-based, Backbone=BEiT-B/32, Training Stage=w/o Stage 12026.0573.575.194
ALOHaMethod Category=LLM-based2026.0570.9——
SPECSMethod Category=Encoder-based2026.0569.872.296.2
GPT4o (API)Method Category=LLM-based2026.05——90.9