Share your thoughts, 1 month free Claude Pro on usSee more

Mathematical Reasoning on MATH 500 (Acc, SL, SE)

99.1Accuracy (Acc)

Qwen3-Base SAT

Updated 7d ago

Evaluation Results

Method
Qwen3-Base SAT 2026.04	99.1	-	-	-	-	-
Qwen3-A3B 2026.04	98	-	-	-	-	-
QwQ 2026.04	98	-	-	-	-	-
OpenAI o3-mini (medium) 2026.04	98	-	-	-	-	-
Qwen3 2026.04	97.8	-	-	-	-	-
I-DLM 2026.04	97.6	-	-	-	-	-
Qwen3 (thinking) 2026.04	97.2	-	-	-	-	-
I-DLM 2026.04	96.8	-	-	-	-	-
KnowRL-Nemotron-1.5B 2026.04	96.45	-	-	-	-	-
KnowRL-Nemotron-1.5B 2026.04	96.2	-	-	-	-	-
Qwen3-Base Debate + Judge 2026.04	95.9	-	-	-	-	-
Qwen3 2026.04	95.8	-	-	-	-	-
KnowRL-Nemotron-1.5B 2026.04	95.7	-	-	-	-	-
Qwen3-Base Role-play 2026.04	95.7	-	-	-	-	-
StepFlow 2026.04	95.6	-	-	-	-	-
Qwen3-Base Debate 2026.04	95.2	-	-	-	-	-
LessIsMore 2025.08	95.14	-	-	-	-	-
LessIsMore 2025.08	95.12	-	-	-	-	-
Full Attn 2025.08	95.1	-	-	-	-	-
JustRL 2026.04	94.85	-	-	-	-	-
LessIsMore 2025.08	94.85	-	-	-	-	-
JustRL 2026.04	94.59	-	-	-	-	-
LessIsMore 2025.08	94.55	-	-	-	-	-
DeepSeek-R1 Distill Llama 2026.04	94.5	-	-	-	-	-
LessIsMore 2025.08	94.45	-	-	-	-	-
Full Attn 2025.08	94.43	-	-	-	-	-
LessIsMore 2025.08	94.42	-	-	-	-	-
LessIsMore 2025.08	94.4	-	-	-	-	-
DeepSeek-R1-Distill-Qwen 32B 2026.04	94.3	-	-	-	-	-
LessIsMore 2025.08	94.22	-	-	-	-	-
LessIsMore 2025.08	94.16	-	-	-	-	-
JustRL 2026.04	94.15	-	-	-	-	-
LessIsMore 2025.08	94.12	-	-	-	-	-
QuestA 2026.04	94.05	-	-	-	-	-
QuestA 2026.04	93.94	-	-	-	-	-
Full Attn 2025.08	93.93	-	-	-	-	-
Pass@k† 2026.05	93.67	-	-	-	-	18,788
LessIsMore 2025.08	93.35	-	-	-	-	-
QuestA 2026.04	92.95	-	-	-	-	-
Nemotron-1.5B 2026.04	92.9	-	-	-	-	-
Nemotron-1.5B 2026.04	92.65	-	-	-	-	-
LessIsMore 2025.08	92.5	-	-	-	-	-
HAPO 2025.09	92.47	-	-	-	-	-
Nemotron-1.5B 2026.04	92.35	-	-	-	-	-
ASLEC-CASL 2026.04	92.2	-	-	-	-	-
Original 2025.08	92.2	4,223	-	-	-	-
StepFlow 2026.04	92	-	-	-	-	-
Qwen3-Base-DAPO 2026.04	91.5	-	-	-	-	-
LCPO 2025.08	91.4	2,033	-	-0.8	-51.86	-
Pass@k† 2026.05	91.07	-	-	-	-	16,774
CE-GPPO 2025.09	91	-	-	-	-	-
Qwen3-Base-GRPO 2026.04	90.7	-	-	-	-	-
QwQ-32B-Preview 2026.04	90.6	-	-	-	-	-
ASLEC-DROP 2026.04	90.6	-	-	-	-	-
AdaptThink 2025.08	90.6	1,975	-	-1.6	-53.23	-
GSPO 2025.09	90.3	-	-	-	-	-
RL-PLUS 2025.07	90.2	-	-	-	-	-
RL-PLUS 2025.07	90.2	-	-	-	-	-
Qwen3-Base 2026.04	90.1	-	-	-	-	-
Re-Schedule 2025.10	89.8	-	-	-	-	-
AutoThink 2025.08	89.4	2,187	-	-2.8	-48.21	-
L1 2026.03	89.2	2,603	3.43	-	-	-
EAT 2026.03	89.2	4,297	2.08	-	-	-
EDGE-GRPO 2025.09	89.12	-	-	-	-	-
JET 2025.09	89	2,922	-	-	-	-
DAPO w/ Forking Tokens 2025.09	88.9	-	-	-	-	-
ACC 2025.10	88.8	-	-	-	-	-
LessIsMore 2025.08	88.75	-	-	-	-	-
Qwen3-Base 2026.04	88.6	-	-	-	-	-
LessIsMore 2025.08	88.54	-	-	-	-	-
Entropy Env 2025.09	88.35	-	-	-	-	-
Vanilla 2026.03	88.33	4,554	1.94	-	-	-
Vanilla DAPO 2025.09	88.25	-	-	-	-	-
LessIsMore 2025.08	88.15	-	-	-	-	-
Base 2025.09	88	4,299	-	-	-	-
SPS 2026.05	87.87	-	-	-	-	9,076
Archer 2025.09	87.76	-	-	-	-	-
LUFFY 2025.07	87.6	-	-	-	-	-
TIES-Merging* (SFT + On-Policy RL) 2026.05	87.6	-	-	-	-	-
ROM 2026.03	87.5	3,063	2.86	-	-	-
ROM_CSC 2026.03	87.5	2,807	3.12	-	-	-
PB-SMC 2026.05	87.47	-	-	-	-	10,628
DoTS (ExGRPO + ReLIFT) 2026.05	87.4	-	-	-	-	-
SFT+GRPO 2025.07	87.2	-	-	-	-	-
SFT+RL 2026.05	87.2	-	-	-	-	-
LessIsMore 2025.08	86.9	-	-	-	-	-
ReLIFT 2025.07	86.8	-	-	-	-	-
LUFFY 2026.05	86.8	-	-	-	-	-
DoTS (SFT + On-Policy RL) 2026.05	86.8	-	-	-	-	-
DVTS 2026.05	86.73	-	-	-	-	22,978
SMC 2026.05	86.67	-	-	-	-	23,307
DAPO 2025.09	86.6	3,258	-	-	-	-
ASLEC-CASL 2026.04	86.6	-	-	-	-	-
Qwen3-14B w. KL-Cov 2026.04	86.4	-	-	-	-	-
TEPO 2026.04	86.4	-	-	-	-	-
Best-of-N 2026.05	86.4	-	-	-	-	18,788
Beam Search 2026.05	86.27	-	-	-	-	22,771
ASLEC-DROP 2026.04	86.2	-	-	-	-	-
ReLIFT 2026.05	86.2	-	-	-	-	-
CoT-SC 2025.10	86.2	-	-	-	-	-

Showing 100 of 545 rows