List your company

Code on MBPP 1,000-example (test)

9.0212PerplexityQwen3-VL-2B-Instruct
8.5238811.8807915.237718.59461Jan 13, 2026

Evaluation results

MethodLinks
Qwen3-VL-2B-InstructAblation State=SRA, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.019.0212-0.2403
Qwen3-VL-2B-InstructAblation State=Base, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.019.2615—
Ministral-14B-Instruct-2512Ablation State=SRA, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.0111.1405-0.3506
Ministral-14B-Instruct-2512Ablation State=Base, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.0111.4911—
Qwen3-VL-8B-InstructAblation State=SRA, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.0111.8694-0.2965
Qwen3-VL-8B-InstructAblation State=Base, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.0112.1659—
Qwen3-VL-4B-InstructAblation State=SRA, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.0113.5447-0.2597
Qwen3-VL-4B-InstructAblation State=Base, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.0113.8044—
Ministral-3B-Instruct-2512Ablation State=SRA, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.0120.3751-1.0791
Ministral-3B-Instruct-2512Ablation State=Base, Evaluation Mode=Teacher-forced, Sequence Length=Full-sequence, Seed=422026.0121.4542—