List your company

Mathematical Reasoning on AMC (ACC, LEN)

100Accuracy (ACC)SAT
71.478.82586.2593.675Sep 27, 2025Oct 29, 2025Dec 1, 2025Jan 2, 2026Feb 4, 2026Mar 8, 2026Apr 10, 2026

Evaluation results

MethodLinks
SATModel=Qwen3-14B2026.041004,848
SATModel=Qwen3-14B2026.041004,848
DEERModel=Qwen3-32B2026.0497.55,753
SATModel=Qwen3-32B2026.0497.54,953
DEERModel=Qwen3-32B2026.0497.55,753
SATModel=Qwen3-32B2026.0497.54,953
DEERModel=Qwen3-14B2026.04954,763
COTModel=Qwen3-32B2026.04956,448
SATModel=Nemo-Llama-8b2026.04954,441
DEERModel=QwQ-32B2026.04955,782
SATModel=QwQ-32B2026.04955,777
DEERModel=Qwen3-14B2026.04954,763
COTModel=Qwen3-32B2026.04956,448
SATModel=Nemo-Llama-8b2026.04954,441
DEERModel=QwQ-32B2026.04955,782
SATModel=QwQ-32B2026.04955,777
CGRSModel=Qwen3-32B2026.0494.24,766
CGRSModel=Qwen3-32B2026.0494.24,766
CGRSModel=Qwen3-14B2026.0493.35,076
TALEModel=Qwen3-32B2026.0493.36,533
CGRSModel=QwQ-32B2026.0493.34,771
CGRSModel=Qwen3-14B2026.0493.35,076
CGRSModel=QwQ-32B2026.0493.34,771
COTModel=Qwen3-4B2026.0492.57,408
SATModel=Qwen3-4B2026.0492.55,255
DEERModel=Qwen3-8B2026.0492.54,392
SATModel=Qwen3-8B2026.0492.55,308
COTModel=Qwen3-14B2026.0492.56,964
TALEModel=Qwen3-14B2026.0492.55,951
DynasorModel=Qwen3-32B2026.0492.55,521
SATModel=DS-Qwen-7B2026.0492.53,770
COTModel=DS-Qwen-14B2026.0492.55,333
SATModel=DS-Qwen-14B2026.0492.54,656
COTModel=Qwen3-4B2026.0492.57,408
SATModel=Qwen3-4B2026.0492.55,255
DEERModel=Qwen3-8B2026.0492.54,392
SATModel=Qwen3-8B2026.0492.55,308
COTModel=Qwen3-14B2026.0492.56,964
SATModel=DS-Qwen-7B2026.0492.53,770
COTModel=DS-Qwen-14B2026.0492.55,333
SATModel=DS-Qwen-14B2026.0492.54,656
TALEModel=QwQ-32B2026.0490.86,522
VanillaModel=Qwen3-8B, Budget=24002026.0390—
VanillaModel=Qwen3-8B, Budget=32002026.0390—
COTModel=Qwen3-8B2026.04907,920
DynasorModel=Qwen3-14B2026.04906,030
COTModel=Nemo-Llama-8b2026.04905,997
DEERModel=Nemo-Llama-8b2026.04905,408
COTModel=QwQ-32B2026.04907,086
DynasorModel=QwQ-32B2026.04907,185
COTModel=Qwen3-8B2026.04907,920
COTModel=Nemo-Llama-8b2026.04905,997
DEERModel=Nemo-Llama-8b2026.04905,408
COTModel=QwQ-32B2026.04907,086
DynasorModel=Qwen3-8B2026.0489.26,457
CGRSModel=Qwen3-8B2026.0489.25,595
CGRSModel=Qwen3-8B2026.0489.25,595
TALEModel=Qwen3-8B2026.0488.36,872
CGRSModel=DS-Qwen-7B2026.0488.33,406
NoThinkingModel=QwQ-32B2026.0488.37,493
CGRSModel=DS-Qwen-7B2026.0488.33,406
DEERModel=Qwen3-4B2026.0487.54,906
COTModel=DS-Qwen-7B2026.0487.56,366
DEERModel=Qwen3-4B2026.0487.54,906
COTModel=DS-Qwen-7B2026.0487.56,366
TALEModel=Qwen3-4B2026.0486.75,107
DynasorModel=Qwen3-4B2026.0486.76,233
CGRSModel=Qwen3-4B2026.0486.74,351
TALEModel=DS-Qwen-7B2026.0486.75,107
CGRSModel=Qwen3-4B2026.0486.74,351
DYPOBase Model=Qwen3-4B-Base2026.0486—
R-KVModel=Qwen3-8B, Budget=24002026.0385—
LongFlowModel=Qwen3-8B, Budget=32002026.0385—
DEERModel=DS-Qwen-7B2026.04854,451
DEERModel=DS-Qwen-14B2026.04854,240
DEERModel=DS-Qwen-7B2026.04854,451
DEERModel=DS-Qwen-14B2026.04854,240
DynasorModel=DS-Qwen-7B2026.0484.25,201
LongFlowModel=DeepSeek-R1-Distill-Llama-8B, Budget=32002026.0382.5—
R-KVModel=Qwen3-8B, Budget=32002026.0382.5—
VATPModel=Qwen3-8B, Budget=32002026.0382.5—
LongFlowModel=Qwen3-8B, Budget=24002026.0380—
H2OModel=Qwen3-8B, Budget=32002026.0380—
VanillaModel=DeepSeek-R1-Distill-Llama-8B, Budget=24002026.0377.5—
VanillaModel=DeepSeek-R1-Distill-Llama-8B, Budget=32002026.0377.5—
R-KVModel=DeepSeek-R1-Distill-Llama-8B, Budget=32002026.0377.5—
VATPModel=Qwen3-8B, Budget=24002026.0377.5—
NoThinkingModel=Qwen3-14B2026.0477.51,616
DAPOBackbone=DS-Distill-Llama-8B2025.0977.45,463
BaseBackbone=DS-Distill-Llama-8B2025.0976.97,521
JETBackbone=DS-Distill-Llama-8B2025.0976.85,261
DPOBackbone=DS-Distill-Llama-8B2025.0976.46,450
NoThinkingModel=DS-Qwen-7B2026.0475.82,499
SFT -> RLBase Model=Qwen3-4B-Base2026.0475.4—
R-KVModel=DeepSeek-R1-Distill-Llama-8B, Budget=24002026.0375—
NoThinkingModel=Qwen3-32B2026.04752,221
SFTBackbone=DS-Distill-Llama-8B2025.0972.86,392
VATPModel=DeepSeek-R1-Distill-Llama-8B, Budget=24002026.0372.5—
H2OModel=DeepSeek-R1-Distill-Llama-8B, Budget=32002026.0372.5—
VATPModel=DeepSeek-R1-Distill-Llama-8B, Budget=32002026.0372.5—