Share your thoughts, 1 month free Claude Pro on usSee more

Hallucination Assessment on MHumanEval

72.6Response Rate

LLaVA-RLHF

Updated 1mo ago

Evaluation Results

Method	Links
LLaVA-RLHF 2024.05		72.6
CCA-LLaVA 2024.05		68.5
POVID 2024.05		67.8
VCD 2024.05		67.1
LLaVA 1.5 2024.05		67.1
Less-is-more 2024.05		63.7
OPERA 2024.05		63
Qwen-VL-Chat 2024.05		61
MiniGemini 2024.05		59.6
RLHF-V 2024.05		55.5
AMP-MEG 2024.05		54.8
Silkie 2024.05		54.1
LLaVA-NeXT 2024.05		53.4
HA-DPO 2024.05		53.4
OmniLMM 2024.05		52.7
GPT-4V 2024.05		45.9
LLaVA 1.5 + RLAIF-V 2024.05		44.5
LLaVA 1.5 + RLAIF-V BoN 2024.05		39.7
OmniLMM + RLAIF-V 2024.05		35.6
OmniLMM + RLAIF-V BoN 2024.05		29.5