List your company
ResearchFormerly WizWand

Mathematical Problem Solving on Minerva (accuracy)

31.66AccuracyACPO
12.586417.538222.4927.4418Jun 21, 2026
Updated 4mo ago

Evaluation results

MethodLinks
ACPORegime=OffP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0631.66
ACPORegime=N-OnP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0631.02
CISPORegime=N-OnP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0630.51
DAPORegime=OffP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0630.47
DAPORegime=N-OnP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0629.92
CISPORegime=OffP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0629.55
High-EntropyRegime=OffP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0629.27
AR-LoptiRegime=N-OnP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0629.14
High-EntropyRegime=N-OnP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0629.08
Low-EntropyRegime=N-OnP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0628.72
AR-LoptiRegime=OffP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0625.14
Low-EntropyRegime=OffP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0625.05
Base ModelRegime=N-OnP., Backbone=Qwen2.5-7B, Training Dataset=ORZ-57K, Sampling Strategy=avg@82026.0613.32