List your company

Mathematical Reasoning on AIME (Accuracy (mv %), Mean Questions, Mean Tokens)

86.7Accuracy (mv %)Qwen3-14B
10.36430.1825069.818May 29, 2026

Evaluation results

MethodLinks
Qwen3-14BSetting=Adaptive, Mode=(t)2026.0586.701,133
Qwen3-4BSetting=Adaptive, Mode=(t)2026.058001,300
Qwen3-14BSetting=Baseline, Mode=(t)2026.0580—11,871
Qwen3-4BSetting=Baseline, Mode=(t)2026.0576.7—12,259
Olmo-3-7B-ThinkSetting=Baseline, Mode=(t)2026.0576.7—15,429
Olmo-3-7B-ThinkSetting=Adaptive, Mode=(t)2026.0576.7133,243
Olmo-3-7B-ThinkSetting=Fixed, Mode=(t)2026.0573.3293,340
Qwen3-1.7BSetting=Adaptive, Mode=(t)2026.0553.30.813,744
Qwen3-1.7BSetting=Baseline, Mode=(t)2026.0548.3—13,570
Qwen3-14BSetting=Baseline, Mode=(nt)2026.0543.3—4,324
Qwen3-14BSetting=Fixed, Mode=(nt)2026.05403.117,730
Qwen3-4BSetting=Baseline, Mode=(nt)2026.0536.7—1,271
Qwen3-14BSetting=Adaptive, Mode=(nt)2026.0536.70.93,915
Qwen3-4BSetting=Adaptive, Mode=(nt)2026.053015,317
Qwen3-4BSetting=Fixed, Mode=(nt)2026.0528.62.320,523
Qwen3-1.7BSetting=Baseline, Mode=(nt)2026.0516.7—2,933
Qwen3-1.7BSetting=Adaptive, Mode=(nt)2026.0513.31.24,584
Qwen3-1.7BSetting=Fixed, Mode=(nt)2026.0513.33.816,772