List your company

Reasoning on LiveCodeBench (pass@1 accuracy)

55.45pass@1 AccuracyBaseline
3.7117.142530.57544.0075Oct 1, 2025Nov 13, 2025Dec 26, 2025Feb 7, 2026Mar 22, 2026May 4, 2026Jun 17, 2026

Evaluation results

MethodLinks
BaselineModel=QwQ-32B, Bit-Width=16-162025.1055.45
ThinKV (k=1024)Model=QwQ-32B, Bit-Width=3.4-3.42025.1050.47
BF16 BaselineModel Backbone=Qwen3, Model Size=32B2026.0649.9
GraphPOModel=Deepseek-R1-Distill-Qwen-7B2026.0648.3
BaselineModel=R1-Qwen-14B, Bit-Width=16-162025.1047.9
MR-GPTQModel Backbone=Qwen3, Model Size=14B2026.0647.2
RTNModel Backbone=Qwen3, Model Size=14B2026.0646.7
ReSETModel Backbone=Qwen3, Model Size=32B2026.0646.7
PM-KVQModel=QwQ-32B, Bit-Width=3.5-3.52025.1046.68
BF16 BaselineModel Backbone=Qwen3, Model Size=14B2026.0646.5
RTNModel Backbone=Qwen3, Model Size=32B2026.0646.5
ReSETModel Backbone=Qwen3, Model Size=14B2026.0646.1
ThinKV (k=1024)Model=R1-Qwen-14B, Bit-Width=3.5-3.52025.1045.84
4/6Model Backbone=Qwen3, Model Size=32B2026.0645.7
4/6Model Backbone=Qwen3, Model Size=14B2026.0645.3
BRQModel Backbone=Qwen3, Model Size=32B2026.0645.1
MR-GPTQModel Backbone=Qwen3, Model Size=32B2026.0645.1
PROSModel=Deepseek-R1-Distill-Qwen-7B2026.0644.8
SPOModel=Deepseek-R1-Distill-Qwen-7B2026.0644.7
TREE-GRPOModel=Deepseek-R1-Distill-Qwen-7B2026.0644.1
TreePOModel=Deepseek-R1-Distill-Qwen-7B2026.0643.5
BF16 BaselineModel Backbone=Qwen3, Model Size=8B2026.0643.2
MR-GPTQModel Backbone=Qwen3, Model Size=8B2026.0642.9
TreeRLModel=Deepseek-R1-Distill-Qwen-7B2026.0642.5
ReSETModel Backbone=Qwen3, Model Size=8B2026.0642.1
PM-KVQModel=R1-Qwen-14B, Bit-Width=3.2-3.22025.1041.97
DAPOModel=Deepseek-R1-Distill-Qwen-7B2026.0641.4
BF16 BaselineModel Backbone=All Models, Model Size=Avg2026.0641.3
KIVIModel=QwQ-32B, Bit-Width=2-22025.1040.75
GRPOModel=Deepseek-R1-Distill-Qwen-7B2026.0640.7
ReSETModel Backbone=All Models, Model Size=Avg2026.0640.2
MR-GPTQModel Backbone=All Models, Model Size=Avg2026.0639.5
RTNModel Backbone=All Models, Model Size=Avg2026.0639.2
BRQModel Backbone=Qwen3, Model Size=14B2026.0639
BF16 BaselineModel Backbone=R1-Qwen, Model Size=14B2026.0638.5
4/6Model Backbone=All Models, Model Size=Avg2026.0638
ReSETModel Backbone=R1-Qwen, Model Size=14B2026.0637.9
BRQModel Backbone=R1-Qwen, Model Size=14B2026.0637.7
RTNModel Backbone=R1-Qwen, Model Size=14B2026.0637.1
BRQModel Backbone=All Models, Model Size=Avg2026.0637.1
BaseModel=Deepseek-R1-Distill-Qwen-7B2026.0636.8
RTNModel Backbone=Qwen3, Model Size=8B2026.0636.4
4/6Model Backbone=R1-Qwen, Model Size=14B2026.0636.4
BRQModel Backbone=Qwen3, Model Size=8B2026.0636.2
4/6Model Backbone=Qwen3, Model Size=8B2026.0635.8
MR-GPTQModel Backbone=R1-Qwen, Model Size=14B2026.0634.9
KIVIModel=R1-Qwen-14B, Bit-Width=2-22025.1034.56
GraphPOModel=Qwen3-8B-Base2026.0634.5
PROSModel=Qwen3-8B-Base2026.0631.1
SPOModel=Qwen3-8B-Base2026.0630.8
TREE-GRPOModel=Qwen3-8B-Base2026.0630.4
RTNModel Backbone=R1-Qwen, Model Size=7B2026.0629.5
TreePOModel=Qwen3-8B-Base2026.0629.4
ReSETModel Backbone=R1-Qwen, Model Size=7B2026.0628.4
BF16 BaselineModel Backbone=R1-Qwen, Model Size=7B2026.0628.2
TreeRLModel=Qwen3-8B-Base2026.0628.1
BRQModel Backbone=R1-Qwen, Model Size=7B2026.0627.6
DAPOModel=Qwen3-8B-Base2026.0627.6
MR-GPTQModel Backbone=R1-Qwen, Model Size=7B2026.0627.2
GRPOModel=Qwen3-8B-Base2026.0627.2
4/6Model Backbone=R1-Qwen, Model Size=7B2026.0627
BaseModel=Qwen3-8B-Base2026.0622.9
GraphPOModel=Qwen2.5-7B-Math2026.0616.9
SPOModel=Qwen2.5-7B-Math2026.0613.6
PROSModel=Qwen2.5-7B-Math2026.0613.5
TREE-GRPOModel=Qwen2.5-7B-Math2026.0613.4
TreePOModel=Qwen2.5-7B-Math2026.0613
DAPOModel=Qwen2.5-7B-Math2026.0612.6
TreeRLModel=Qwen2.5-7B-Math2026.0612.5
GRPOModel=Qwen2.5-7B-Math2026.0611.4
BaseModel=Qwen2.5-7B-Math2026.065.7