Knowledge Distillation During Mid-Training Favors Reasoning over Factual Recall
尚未评估计划受阻指标结果claim_t3_sd_baseline

Mid-training ablation results for Switch Distillation absolute macro-averages across Reasoning, Factual Recall, and Knowledge task groups.

来源:source_paper:Table 3, Page 9

报告指标与观测值

macro_average

t3_sd_baseline_reasoning

论文报告 44.7 percentage_points

实际观测 — percentage_points

macro_average

t3_sd_baseline_factual_recall

论文报告 29.3 percentage_points

实际观测 — percentage_points

macro_average

t3_sd_baseline_knowledge

论文报告 49.3 percentage_points

实际观测 — percentage_points

Assessment(0)

这条 Claim 暂无已发布的不可变 Assessment。

关联运行(0)

这条 Claim 暂无关联执行。