List your company
ResearchFormerly WizWand

Mathematical Reasoning on AMC 2024 (Pass@1, Pass@32, Avg)

49.8Pass@1Qwen3-8B-Base
6.53617.7682940.232Sep 28, 2025Oct 23, 2025Nov 17, 2025Dec 12, 2025Jan 6, 2026Jan 31, 2026Feb 26, 2026
Updated 6mo ago

Evaluation results

MethodLinks
Qwen3-8B-BaseTraining=RLAD (ours), Context=30K2026.0249.877.466.5
Qwen3-8B-BaseTraining=GRPO, Context=30K2026.0249.276.961
Qwen3-8B-BaseTraining=KDRL, Context=30K2026.0249.17764.9
Qwen3-8B-BaseTraining=SFT, Context=30K2026.0248.272.956
Qwen2.5-7B + PPOBase Model=Qwen2.5-7B, RL Framework=PPO, VERL Integration=false2025.0933.3——
Qwen2.5-7B + PPO w/ VERL.Base Model=Qwen2.5-7B, RL Framework=PPO, VERL Integration=true2025.0933.3——
Qwen2.5-7B + GRPO w/ VERL.Base Model=Qwen2.5-7B, RL Framework=GRPO, VERL Integration=true2025.0928.9——
Qwen2.5-7B + GRPOBase Model=Qwen2.5-7B, RL Framework=GRPO, VERL Integration=false2025.0926.7——
Qwen3-1.7B-BaseTraining=RLAD (ours), Context=30K2026.0224.462.439
Qwen3-1.7B-BaseTraining=KDRL, Context=30K2026.022159.737.2
Qwen3-1.7B-BaseTraining=GRPO, Context=30K2026.0219.758.636.5
Qwen3-8B-BaseTraining=-, Context=30K2026.0218.864.636.8
Qwen2.5-7BBase Model=Qwen2.5-7B, RL Framework=None, VERL Integration=false2025.0915.6——
Llama-3.2-3B-Instruct + PPOBase Model=Llama-3.2-3B-Instruct, RL Framework=PPO, VERL Integration=false2025.0913.3——
Llama-3.2-3B-InstructBase Model=Llama-3.2-3B-Instruct, RL Framework=None, VERL Integration=false2025.0911.1——
Llama-3.2-3B-Instruct + GRPO w/ VERL.Base Model=Llama-3.2-3B-Instruct, RL Framework=GRPO, VERL Integration=true2025.0911.1——
Llama-3.2-3B-Instruct + PPO w/ VERL.Base Model=Llama-3.2-3B-Instruct, RL Framework=PPO, VERL Integration=true2025.0911.1——
Llama-3.2-3B-Instruct + GRPOBase Model=Llama-3.2-3B-Instruct, RL Framework=GRPO, VERL Integration=false2025.098.9——
Qwen3-1.7B-BaseTraining=-, Context=30K2026.028.244.622.2