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

Mathematical Reasoning on MATH 500 (ACC, LEN, delta_LEN)

99.1Accuracy

Qwen3-Base (DICE-PC)

Updated 29d ago

Evaluation Results

Method
Qwen3-Base (DICE-PC) 2026.06	99.1	-	-	-	-
OpenAI o3 2026.06	98.9	-	-	-	-
Qwen3-235B-A22B 2026.06	98.8	-	-	-	-
Qwen3-Base (DICE-FT) 2026.06	98.6	-	-	-	-
Qwen3-A3B 2026.06	98	-	-	-	-
QwQ 2026.06	98	-	-	-	-
OpenAI o3-mini 2026.06	98	-	-	-	-
Qwen3-Inst (SC@64) 2026.06	97.9	-	-	-	-
Qwen3 (thinking) 2026.06	97.2	-	-	-	-
Qwen3-Inst (Few-shot CoT) 2026.06	96.8	-	-	-	-
Qwen3-Base Debate+Judge (FT) 2026.06	95.9	-	-	-	-
Qwen3-Base Role-play (FT) 2026.06	95.7	-	-	-	-
SRQ-driven steering 2026.04	95.6	-	-	-	-
Qwen3-Inst Debate+Judge (prompt) 2026.06	95.3	-	-	-	-
SRQ-driven steering 2026.04	95.2	-	-	-	-
Qwen3-Base Debate+Judge (FT) 2026.06	95.2	-	-	-	-
Qwen3-Base Debate (FT) 2026.06	95.2	-	-	-	-
Qwen3-Base Role-play (FT) 2026.06	95	-	-	-	-
SRQ-driven steering 2026.04	94.8	-	-	-	-
Qwen3-Inst Role-play (prompt) 2026.06	94.7	-	-	-	-
Conceptor 2026.04	94.6	-	-	-	-
Qwen3-Base Debate (FT) 2026.06	94.6	-	-	-	-
DeepSeek-R1 Distill 2026.06	94.5	-	-	-	-
PCA-CAA 2026.04	94.4	-	-	-	-
Base LLM 2026.04	94.2	-	-	-	-
CAA 2026.04	94.2	-	-	-	-
Qwen3-Inst Debate (prompt) 2026.06	94.1	-	-	-	-
SLAT 2026.05	92.8	1,820	-	-	-
Qwen3-Base-DAPO 2026.06	92.7	-	-	-	-
DAST 2026.05	92.4	3,228	-	-	-
Qwen3-Base-GRPO 2026.06	92.2	-	-	-	-
FT 2026.06	92.18	-	-	-	-
STARE-C2 2026.06	92.1	-	-	-	-
Original 2026.05	91.9	3,860	-	-	-
JustRL 2026.06	91.7	-	-	-	-
AdaptThink 2026.05	91.6	1,893	-	-	-
Qwen3-Base-DAPO 2026.06	91.5	-	-	-	-
A3PO 2026.06	91.3	-	-	-	-
TLMRE 2026.05	91.2	2,662	-	-	-
STARE-C2 2026.06	91.2	-	-	-	-
EAPO 2026.06	91.1	-	-	-	-
ReQAT_TQ 2026.06	90.72	-	-	-	-
TLMRE 2026.05	90.7	2,312	-	-	-
Qwen3-Base-GRPO 2026.06	90.7	-	-	-	-
STARE-C2 2026.06	90.6	-	-	-	-
STARE-O1 2026.06	90.6	-	-	-	-
ReQAT_TQS 2026.06	90.53	-	-	-	-
SRQ-driven steering 2026.04	90.4	-	-	-	-
SRQ-driven steering 2026.04	90.4	-	-	-	-
L1-Max 2026.05	90.4	2,134	-	-	-
STARE-O1 2026.06	90.4	-	-	-	-
STARE-O1 2026.06	90.4	-	-	-	-
STAPO 2026.06	90.4	-	-	-	-
EntroAdv 2026.06	90.2	-	-	-	-
Qwen3-Inst (ReAct) 2026.06	90.1	-	-	-	-
Qwen3-Base-PPO 2026.06	90.1	-	-	-	-
Qwen3-Base 2026.06	90.1	-	-	-	-
SRQ-driven steering 2026.04	90	-	-	-	-
Baseline 2026.06	90	-	-	-	-
Lp-Reg 2026.06	90	-	-	-	-
KL-Cov 2026.06	90	-	-	-	-
DAPO 2026.06	89.9	-	-	-	-
80/20 Rule 2026.06	89.9	-	-	-	-
PCA-CAA 2026.04	89.8	-	-	-	-
LC-R1 2026.05	89.8	1,583	-	-	-
ReQAT_T 2026.06	89.8	-	-	-	-
Conceptor 2026.04	89.6	-	-	-	-
SWA8k-RL-900 2026.06	89.4	-	-	-	-
SWA4k-RL-1400 2026.06	89.4	-	-	-	-
80/20 Rule 2026.06	89.4	-	-	-	-
SWA8k-RL-1200 2026.06	89.3	-	-	-	-
CAA 2026.04	89.2	-	-	-	-
SA-RL-900 2026.06	89.1	-	-	-	-
GRPO 2026.06	88.9	-	-	-	-
GRPO 2026.06	88.9	-	-	-	-
Base LLM 2026.04	88.6	-	-	-	-
Qwen3-Base 2026.06	88.6	-	-	-	-
SWA4k-RL-900 2026.06	88.6	-	-	-	-
GRPO-ds 2026.06	88.6	-	-	-	-
FT + PTQ 2026.06	88.53	-	-	-	-
SWA2k-RL-1700 2026.06	88.5	-	-	-	-
SimpleTIR 2026.06	88.4	-	-	-	-
GRPO-ds 2026.06	88.3	-	-	-	-
Qwen3-Inst (SC@64) 2026.06	88.2	-	-	-	-
SRQ-driven steering 2026.04	87.9	-	-	-	-
Qwen3-Inst (ToT) 2026.06	87.9	-	-	-	-
GSPO 2026.06	87.6	-	-	-	-
DAPO 2026.06	87.6	-	-	-	-
W-REINF 2026.06	87.5	-	-	-	-
SRQ-driven steering 2026.04	87.4	-	-	-	-
Lp-Reg 2026.06	87.4	-	-	-	-
STARE-C2 2026.06	87.2	-	-	-	-
PCA-CAA 2026.04	87.1	-	-	-	-
SWA2k-RL-900 2026.06	87.1	-	-	-	-
SRQ-driven steering 2026.04	87	-	-	-	-
EntroReg 2026.06	86.8	-	-	-	-
STARE-O1 2026.06	86.8	-	-	-	-
GRPO 2026.06	86.6	-	-	-	-
Conceptor 2026.04	86.5	-	-	-	-
Effective TIR 2026.06	86.4	-	-	-	-

Showing 100 of 601 rows