Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

SLR-Bench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Logical ReasoningSLR-BENCH Extended Leaderboard
LRL Score15.5
54
Logical ReasoningSLR-BENCH (test)
LRL11.3
27
Logical ReasoningSLR-BENCH
Overall LRL Score15.5
14
inductive Prolog rule synthesisSLR-Bench Hard tier 250 tasks 1
Accuracy58.4
13
inductive Prolog rule synthesisSLR-Bench Medium tier 250 tasks 1
Accuracy88.8
13
inductive Prolog rule synthesisSLR-Bench Easy tier 1 (250 tasks)
Accuracy100
13
inductive Prolog rule synthesisSLR-Bench Basic tier 250 tasks 1
Accuracy100
13
inductive Prolog rule synthesisSLR-Bench Overall 1,000 tasks (full)
Accuracy (%)86.7
13
Showing 8 of 8 rows