List your company

Mathematical Reasoning on AIME24, AIME25, AMC23, MATH, Olympiad (Aggregate)

47.3Average ScoreGRPO + coupled rhythm credit
36.69239.44642.244.954Oct 15, 2025

Evaluation results

MethodLinks
GRPO + coupled rhythm creditBackbone=Qwen3-4B-Base, Context Length=8K2025.1047.3
GRPO + local-chunk creditBackbone=Qwen3-4B-Base, Context Length=8K2025.1047.2
GRPO + global-anchor creditBackbone=Qwen3-4B-Base, Context Length=8K2025.1047.1
GRPOBackbone=Qwen3-4B-Base, Context Length=8K2025.1044.8
GRPO + random creditBackbone=Qwen3-4B-Base, Context Length=8K2025.1044.7
GRPO + high-entropy creditBackbone=Qwen3-4B-Base, Context Length=8K2025.1044.5
GRPO + coupled rhythm creditBackbone=Qwen3-8B-Base, Context Length=1K2025.1043.2
GRPO + global-anchor creditBackbone=Qwen3-8B-Base, Context Length=1K2025.1042.1
GRPO + local-chunk creditBackbone=Qwen3-8B-Base, Context Length=1K2025.1041.3
GRPO + high-entropy creditBackbone=Qwen3-8B-Base, Context Length=1K2025.1040.1
GRPO + random creditBackbone=Qwen3-8B-Base, Context Length=1K2025.1039.6
GRPOBackbone=Qwen3-8B-Base, Context Length=1K2025.1039.4
GRPO + coupled rhythm creditBackbone=Qwen3-4B-Base, Context Length=1K2025.1039.2
GRPO + local-chunk creditBackbone=Qwen3-4B-Base, Context Length=1K2025.1038.6
GRPO + global-anchor creditBackbone=Qwen3-4B-Base, Context Length=1K2025.1038.2
GRPO + high-entropy creditBackbone=Qwen3-4B-Base, Context Length=1K2025.1037.2
GRPOBackbone=Qwen3-4B-Base, Context Length=1K2025.1037.1
GRPO + random creditBackbone=Qwen3-4B-Base, Context Length=1K2025.1037.1