LTL-guided Reinforcement Learning on Letter Finite-horizon (test)
100Success Rate (SR)SEMLTL
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SEMLTLTask Specification Type=FINITE-REACTIVE[8, 2], |Q|=300, |δ|=2, 1152026.02 | 100 | 2.8 | |
| SEMLTLTask Specification Type=LOCAL-SAFETY[3, 3], |Q|=233, |δ|=1, 3722026.02 | 99 | 5.38 | |
| DEEPLTLTask Specification Type=SMALL, |Q|=9, |δ|=232026.02 | 98 | — | |
| SEMLTLTask Specification Type=SMALL, |Q|=9, |δ|=232026.02 | 98 | 4.12 | |
| SEMLTLTask Specification Type=GLOBAL-SAFETY[4, 6], |Q|=1, 072, |δ|=6, 4142026.02 | 91 | 6.77 | |
| LTL2ACTIONTask Specification Type=FINITE-REACTIVE[8, 2], |Q|=300, |δ|=2, 1152026.02 | 74 | — | |
| LTL2ACTIONTask Specification Type=SMALL, |Q|=9, |δ|=232026.02 | 67 | — | |
| LTL2ACTIONTask Specification Type=LOCAL-SAFETY[3, 3], |Q|=233, |δ|=1, 3722026.02 | 57 | — | |
| LTL2ACTIONTask Specification Type=GLOBAL-SAFETY[4, 6], |Q|=1, 072, |δ|=6, 4142026.02 | 12 | — |