Tasks SotA Datasets Papers Methods Submit About

Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable Benchmarks All SotA Datasets Papers Methods

Community

Submit Results About

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/BT-Adapter (zero-shot)

BT-Adapter (zero-shot)

Reported on 35 benchmarks across 6 tasks · 1 paper · 1 SOTA

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

Reasoning34 results

Video Question AnsweringonActivityNet-QA
Accuracy· 2023-09-27
46.1
best: 61.6 (Tarsier (34B))
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonActivityNet-QA
Confidence score· 2023-09-27
3.6
best: 2.2 (Video Chat)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonMSVD-QA
Accuracy· 2023-09-27
67
best: 80.3 (Tarsier (34B))
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonMSVD-QA
Confidence Score· 2023-09-27
3.6
best: 2.5 (Video LLaMA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonMSVD-QA
Accuracy· 2023-09-27
67
best: 80.3 (Tarsier (34B))
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonMSVD-QA
Confidence Score· 2023-09-27
3.6
best: 2.5 (Video LLaMA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonMSRVTT-QA
Accuracy· 2023-09-27
51.2
best: 72.4 (Flash-VStream)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonMSRVTT-QA
Confidence Score· 2023-09-27
2.9
best: 1.8 (Video LLaMA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonMSRVTT-QA
Accuracy· 2023-09-27
51.2
best: 72.4 (Flash-VStream)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonMSRVTT-QA
Confidence Score· 2023-09-27
2.9
best: 1.8 (Video LLaMA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonActivityNet-QA
Accuracy· 2023-09-27
46.1
best: 61.6 (Tarsier (34B))
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video Question AnsweringonActivityNet-QA
Confidence Score· 2023-09-27
3.2
best: 1.1 (Video LLaMA)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
Consistency· 2023-09-27
2.2
best: 3.81 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
Contextual Understanding· 2023-09-27
2.89
best: 4.21 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
Correctness of Information· 2023-09-27
2.16
best: 3.85 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
Detail Orientation· 2023-09-27
2.46
best: 3.56 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
Temporal Understanding· 2023-09-27
2.13
best: 3.23 (VLM-RLAIF)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
mean· 2023-09-27
2.46
best: 3.73 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
gpt-score· 2023-09-27
2.89
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
gpt-score· 2023-09-27
2.16
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
gpt-score· 2023-09-27
2.46
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
gpt-score· 2023-09-27
2.13
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Generative Visual Question AnsweringonVideoInstruct
gpt-score· 2023-09-27
2.2
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
Consistency· 2023-09-27
2.2
best: 3.81 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
Contextual Understanding· 2023-09-27
2.89
best: 4.21 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
Correctness of Information· 2023-09-27
2.16
best: 3.85 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
Detail Orientation· 2023-09-27
2.46
best: 3.56 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
Temporal Understanding· 2023-09-27
2.13
best: 3.23 (VLM-RLAIF)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
mean· 2023-09-27
2.46
best: 3.73 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
gpt-score· 2023-09-27
2.89
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
gpt-score· 2023-09-27
2.16
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
gpt-score· 2023-09-27
2.46
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
gpt-score· 2023-09-27
2.13
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Video-based Generative Performance BenchmarkingonVideoInstruct
gpt-score· 2023-09-27
2.2
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785

Natural Language Processing21 results

Question AnsweringonActivityNet-QA
Accuracy· 2023-09-27
46.1
best: 61.6 (Tarsier (34B))
SOTA
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Question AnsweringonMSVD-QA
Accuracy· 2023-09-27
67
best: 80.3 (Tarsier (34B))
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Question AnsweringonMSVD-QA
Confidence Score· 2023-09-27
3.6
best: 2.5 (Video LLaMA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Question AnsweringonMSVD-QA
Accuracy· 2023-09-27
67
best: 80.3 (Tarsier (34B))
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Question AnsweringonMSVD-QA
Confidence Score· 2023-09-27
3.6
best: 2.5 (Video LLaMA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Question AnsweringonMSRVTT-QA
Accuracy· 2023-09-27
51.2
best: 72.4 (Flash-VStream)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Question AnsweringonMSRVTT-QA
Confidence Score· 2023-09-27
2.9
best: 1.8 (Video LLaMA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Question AnsweringonMSRVTT-QA
Accuracy· 2023-09-27
51.2
best: 72.4 (Flash-VStream)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Question AnsweringonMSRVTT-QA
Confidence Score· 2023-09-27
2.9
best: 1.8 (Video LLaMA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Question AnsweringonActivityNet-QA
Confidence Score· 2023-09-27
3.2
best: 1.1 (Video LLaMA)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
Consistency· 2023-09-27
2.2
best: 3.81 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
Contextual Understanding· 2023-09-27
2.89
best: 4.21 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
Correctness of Information· 2023-09-27
2.16
best: 3.85 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
Detail Orientation· 2023-09-27
2.46
best: 3.56 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
Temporal Understanding· 2023-09-27
2.13
best: 3.23 (VLM-RLAIF)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
mean· 2023-09-27
2.46
best: 3.73 (PPLLaVA-7B-dpo)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
gpt-score· 2023-09-27
2.89
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
gpt-score· 2023-09-27
2.16
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
gpt-score· 2023-09-27
2.46
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
gpt-score· 2023-09-27
2.13
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785
Visual Question Answering (VQA)onVideoInstruct
gpt-score· 2023-09-27
2.2
best: 4.21 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785

Other1 result

Video-based Generative Performance Benchmarking (Correctness of Information)onVideoInstruct
gpt-score· 2023-09-27
2.16
best: 3.85 (PPLLaVA-7B)
BT-Adapter: Video Conversation is Feasible Without Video Instruction Tuning arXiv:2309.15785