RLHF-V

Reported on 4 benchmarks across 3 tasks · 1 paper · 4 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing4 results

Visual Question Answering (VQA)onVQA v2
Accuracy· 2023-12-01
80
SOTA
RLHF-V: Towards Trustworthy MLLMs via Behavior Alignment from Fine-grained Correctional Human Feedback arXiv:2312.00849
Image CaptioningonObject HalBench
chair_i· 2023-12-01
7.5
SOTA
RLHF-V: Towards Trustworthy MLLMs via Behavior Alignment from Fine-grained Correctional Human Feedback arXiv:2312.00849
Image CaptioningonObject HalBench
chair_s· 2023-12-01
12.2
SOTA
RLHF-V: Towards Trustworthy MLLMs via Behavior Alignment from Fine-grained Correctional Human Feedback arXiv:2312.00849
Visual Question AnsweringonVQA v2
Accuracy· 2023-12-01
80
SOTA
RLHF-V: Towards Trustworthy MLLMs via Behavior Alignment from Fine-grained Correctional Human Feedback arXiv:2312.00849