Video Question Answering on IntentQA (original test)
66.1Temporal Accuracyvideo-grounded entailment tree reasoning
Evaluation results
| Method | Links | ||
|---|---|---|---|
| video-grounded entailment tree reasoningReasoner=VideoLLAVA (7B)2025.01 | 66.1 | 66.4 | |
| LLoViReasoner=GPT-4 (1.8T)2025.01 | 65.5 | 68.7 | |
| VideoAgentReasoner=GPT-4 (1.8T)2025.01 | 64.1 | 66.5 | |
| VideoTreeReasoner=GPT-4 (1.8T)2025.01 | 61.9 | 66.1 |