Across three independent inference seeds with the trained checkpoint and evaluation settings fixed, JAM’s sample standard deviation is below 1.5 points on all four benchmarks (range 0.31–1.34). Variability is lower on LongMemEval and LoCoMo than on NarrativeQA and HotpotQA.

Across three independent inference seeds with the trained checkpoint and evaluation settings fixed, JAM’s sample standard deviation is below 1.5 points on all four benchmarks (range 0.31–1.34). Variability is lower on LongMemEval and LoCoMo than on NarrativeQA and HotpotQA. · CiteArk

测量结果

已评估 0/20 项

获得支持 0有争议或冲突 0尚未评估 20

保留同一结论下的完整结果组。已有评估排在前面；单项判断不替代整组结论。展开行查看条件、来源与历次证据。

每行是一项测量，分别显示论文报告、最近可用实测和累计证据状态。
指标与条件	论文报告	最近可用实测	证据判断
benchmark: LoCoMo F1 · statisticOrRun: Run 1	52.09 F1 points (0–100)	—	尚未评估
benchmark: LoCoMo F1 · statisticOrRun: Run 2	52.68 F1 points (0–100)	—	尚未评估
benchmark: LoCoMo F1 · statisticOrRun: Run 3	51.1 F1 points (0–100)	—	尚未评估
benchmark: LoCoMo F1 · statisticOrRun: Mean	51.96 F1 points (0–100)	—	尚未评估
benchmark: LoCoMo F1 · statisticOrRun: Std.	0.8 F1 points (0–100)	—	尚未评估
benchmark: LongMemEval Acc. · statisticOrRun: Run 1	65.2%	—	尚未评估
benchmark: LongMemEval Acc. · statisticOrRun: Run 2	64.6%	—	尚未评估
benchmark: LongMemEval Acc. · statisticOrRun: Run 3	65%	—	尚未评估
benchmark: LongMemEval Acc. · statisticOrRun: Mean	64.93%	—	尚未评估
benchmark: LongMemEval Acc. · statisticOrRun: Std.	0.31 percentage points	—	尚未评估
benchmark: NarrativeQA F1 · statisticOrRun: Run 1	45.72 F1 points (0–100)	—	尚未评估
benchmark: NarrativeQA F1 · statisticOrRun: Run 2	43.37 F1 points (0–100)	—	尚未评估

Across three independent inference seeds with the trained checkpoint and evaluation settings fixed, JAM’s sample standard deviation is below 1.5 points on all four benchmarks (range 0.31–1.34). Variability is lower on LongMemEval and LoCoMo than on NarrativeQA and HotpotQA.

报告指标与观测值

Assessment（0）

关联运行（0）

测量结果