ResearchTasksTarget/oracle predictionTarget/oracle predictionFollowBenchmarksDatasetSOTA methodSortMost resultsRecently updatedMost papersApplyDatasetSOTA methodMetricTrendResultsLast updatedObserver target oracle Non-endpoint features (Held-out evals)receiver-gated response law91.74Accuracy1Jun 12, 2026
Observer target oracle Non-endpoint features (Held-out evals)receiver-gated response law91.74Accuracy1Jun 12, 2026