ResearchTasksReward Weighting AdaptationFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedCancer (held-out λnk ∈ {1, 2, 3, 4, 6, 7, 8, 9})Kθ − Uˆ102.49Reward8Mar 4, 2026