List your company

Reasoning on Countdown

83.2AccuracyDMPO-LLaDA-1.5
9.67228.76147.8566.939Sep 25, 2025Nov 3, 2025Dec 13, 2025Jan 22, 2026Mar 2, 2026Apr 11, 2026May 21, 2026

Evaluation results

MethodLinks
DMPO-LLaDA-1.5Sequence Length=5122025.1083.2
DMPO-LLaDASequence Length=5122025.1082.81
DMPO-LLaDASequence Length=2562025.1080.86
DMPO-LLaDA-1.5Sequence Length=2562025.1079.3
DMPO-LLaDA-SFTSequence Length=5122025.1077.34
DMPO-LLaDASequence Length=1282025.1067.19
DMPO-LLaDA-SFTSequence Length=2562025.1067.19
DMPO-LLaDA-1.5Sequence Length=1282025.1059.77
StepMerge (d2)Post-training=true2025.0956.6
DMPO-LLaDA-SFTSequence Length=1282025.1054.69
wd1Post-training=true, Evaluation Source=Corresponding paper2025.0951.2
cGRPO-LLaDASequence Length=2562025.1042.58
diffu-GRPO (d1)Post-training=true, Evaluation Source=Corresponding paper2025.0942.2
cGRPO-LLaDASequence Length=5122025.1037.11
d1-LLaDASequence Length=1282025.1034.38
LIFT2Backbone=LLaDA-1.52026.0531.3
LIFT3Backbone=LLaDA-1.52026.0531.2
d1-LLaDASequence Length=5122025.1030.47
cGRPO-LLaDASequence Length=1282025.1030.08
Dream-Instruct (7B)Sequence Length=2562025.1028.52
LIFT2Backbone=LLaDA-8B-Instruct2026.0527.9
Dream-Instruct (7B)Sequence Length=5122025.1027.34
AlfaBackbone=LLaDA-8B-Instruct, LoRA Rank=64, Tuned Params (Usage %)=69.2M (0.85%), Steps=128, Zero-shot=true2026.0327.3
d1-LLaDASequence Length=2562025.1026.56
LIFT3Backbone=LLaDA-8B-Instruct2026.0526.4
LLaDA-1.5 (8B)Sequence Length=1282025.1026.17
LLaDA-1.5 (8B)Sequence Length=5122025.1023.83
LLaDA-Instruct (8B)Sequence Length=1282025.1023.44
LLaDA 1.5Post-training=false, Evaluation Source=Released checkpoint2025.0923.4
CARTBackbone=LLaDA-8B-Instruct2026.0523
Dream-Instruct (7B)Sequence Length=1282025.1022.66
LLaDA 1.5Backbone=LLaDA-1.52026.0522.6
VanillaBackbone=LLaDA-1.52026.0522
GIFTBackbone=LLaDA-8B-Instruct2026.0521.7
DoRA†Backbone=LLaDA-8B-Instruct, LoRA Rank=128, Tuned Params (Usage %)=101.1M (1.25%), Steps=128, Zero-shot=true2026.0321.5
CARTBackbone=LLaDA-1.52026.0521.5
VanillaBackbone=LLaDA-8B-Instruct2026.0520.7
GIFTBackbone=LLaDA-1.52026.0520.7
LoRABackbone=LLaDA-8B-Instruct, LoRA Rank=128, Tuned Params (Usage %)=100.7M (1.24%), Steps=128, Zero-shot=true2026.0320.3
LLaDAPost-training=false, Evaluation Source=Released checkpoint2025.0919.9
LLaDABackbone=LLaDA-8B-Instruct2026.0519.6
AlfaBackbone=LLaDA-8B-Instruct, LoRA Rank=64, Tuned Params (Usage %)=69.2M (0.85%), Steps=256, Zero-shot=true2026.0317.2
LLaDA-1.5 (8B)Sequence Length=2562025.1016.41
LoRA†Backbone=LLaDA-8B-Instruct, LoRA Rank=64, Tuned Params (Usage %)=50.3M (0.62%), Steps=128, Zero-shot=true2026.0316.4
DoRA†Backbone=LLaDA-8B-Instruct, LoRA Rank=128, Tuned Params (Usage %)=101.1M (1.25%), Steps=256, Zero-shot=true2026.0316
LLaDA-Instruct (8B)Sequence Length=5122025.1014.84
LoRABackbone=LLaDA-8B-Instruct, LoRA Rank=128, Tuned Params (Usage %)=100.7M (1.24%), Steps=256, Zero-shot=true2026.0314.5
LLaDA-Instruct (8B)Sequence Length=2562025.1014.45
LoRA†Backbone=LLaDA-8B-Instruct, LoRA Rank=64, Tuned Params (Usage %)=50.3M (0.62%), Steps=256, Zero-shot=true2026.0312.5