Downstream task evaluation after post-training stage RLVR1 for 13b_rkd.
Source: source_paper:Table 8, Page 33, After RLVR1
Reported and observed measurements
downstream_accuracy
t8_rlvr1_13b_rkd_gsm8k
Reported 75.6 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_gsm_s
Reported 60.1 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_gsmplus
Reported 50.3 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_bbh
Reported 33.3 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_drop
Reported 38.1 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_math
Reported 15.8 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_tqa
Reported 53.3 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_nq
Reported 22.3 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_sqa
Reported 8 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_mmlu
Reported 36.2 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_mmlu_p
Reported 16.9 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_arc_c
Reported 55.9 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_obqa
Reported 55.2 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_wino
Reported 51.3 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_agi
Reported 36.4 percentage_points
Observed — percentage_points
downstream_accuracy
t8_rlvr1_13b_rkd_ife
Reported 63.6 percentage_points
Observed — percentage_points
Assessments (0)
No immutable Assessment has been published for this Claim yet.
Runs (0)
No execution has been linked to this Claim yet.