Single-temporal motif recognition on LLMTM standard (test)
1003-star Acctool-augmented LLM agent
Evaluation Results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| tool-augmented LLM agentBackbone=GPT-4o-mini, Method=Agent2025.12 | 100 | 6,689 | 100 | 5,651 | 100 | 5,551 | 100 | 5,685 | 100 | 5,835 | |
| structure-aware dispatcherBackbone=GPT-4o-mini, Method=Ours2025.12 | 88 | 4,165 | 92.5 | 4,002 | 97.5 | 4,141 | 95 | 4,105 | 96 | 4,309 | |
| SingleBackbone=GPT-4o-mini, Method=Single2025.12 | 84.5 | 3,788 | 88.5 | 3,807 | 93 | 3,653 | 93 | 3,776 | 90.5 | 3,832 | |
| RandomBackbone=GPT-4o-mini, Method=Random2025.12 | 80 | 4,430 | 75.5 | 3,893 | 71 | 3,746 | 71 | 3,900 | 72.5 | 3,970 | |
| structure-aware dispatcherBackbone=openPangu-7B, Method=Ours2025.12 | 78 | 5,906 | 77 | 7,042 | 86.5 | 7,248 | 80.5 | 7,117 | 86 | 7,452 | |
| GPT-4o-miniBackbone=GPT-4o-mini, Method=GPT-4o-mini2025.12 | 68.5 | 1,685 | 57.5 | 1,669 | 50 | 1,580 | 53 | 1,630 | 51 | 1,647 |