ResearchTasksTask Vector PerformanceTask Vector PerformanceFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedLlama Baseline 3.2-3BLTV78.65Accuracy6Apr 13, 2026Llama3.2-3B Scenario 5: ICL promptsLTV82.8Accuracy3Apr 13, 2026Llama3.2-3B Scenario 4: More layers & Pos.LTV46.38Accuracy3Apr 13, 2026Llama3.2-3B Scenario 3: More layersLTV80.43Accuracy3Apr 13, 2026Llama Scenario 1: Diff. Pos. 3.2-3BLTV74.1Accuracy3Apr 13, 2026