Human mobility prediction on Saopaulo (test)
23Top-1 AccuracyLLM-Move
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| LLM-Movebase_llm=GPT-4o-mini, training_mode=zero-shot2026.04 | 23 | 37.5 | 0.321 | |
| AgentMovebase_llm=GPT-4o-mini, training_mode=zero-shot2026.04 | 21.5 | 37 | 0.296 | |
| ARMovebase_llm=GPT-4o-mini, training_mode=zero-shot2026.04 | 20 | 39 | 0.3 | |
| LSTPMtraining_mode=supervised2026.04 | 19 | 36.5 | 0.281 | |
| STHGCNtraining_mode=supervised2026.04 | 17.5 | 39.8 | 0.299 | |
| GETNexttraining_mode=supervised2026.04 | 16.5 | 39.8 | 0.299 | |
| DeepMovetraining_mode=supervised2026.04 | 15 | 31 | 0.236 | |
| LLM-ZSbase_llm=GPT-4o-mini, training_mode=zero-shot2026.04 | 15 | 35.5 | 0.257 | |
| LLM-Mobbase_llm=GPT-4o-mini, training_mode=zero-shot2026.04 | 11.5 | 27.5 | 0.2 | |
| RNNtraining_mode=supervised2026.04 | 9.5 | 23 | 0.169 | |
| FPMCtraining_mode=supervised2026.04 | 4.5 | 8.5 | 0.066 |