List your company

Instruction Following on SelfInt (ROUGE-L)

19.42ROUGE-LTeacher (Qwen1.5)
4.1328.10112.0716.039Mar 23, 2026

Evaluation results

MethodLinks
Teacher (Qwen1.5)Distillation Setting=Cross-Tokenizer KD2026.0319.42
Teacher (GPT-2 XL)Distillation Setting=Same-Tokenizer KD2026.0314.65
DSKD-CMA-GADistillation Setting=Cross-Tokenizer KD, Divergence Function=SKL2026.0311.82
DSKD-CMADistillation Setting=Cross-Tokenizer KD, Divergence Function=SRKL2026.0311.5
DSKD-CLADistillation Setting=Cross-Tokenizer KD2026.0311.44
DSKD-CLPDistillation Setting=Cross-Tokenizer KD2026.0311.35
DSKD-CMA-GADistillation Setting=Cross-Tokenizer KD, Divergence Function=KL2026.0311.17
DSKD-CMA-CTDistillation Setting=Cross-Tokenizer KD, Divergence Function=KL2026.0310.83
DSKDDistillation Setting=Same-Tokenizer KD2026.0310.58
MinEDDistillation Setting=Cross-Tokenizer KD2026.039.78
DSKD-CMADistillation Setting=Cross-Tokenizer KD, Divergence Function=KL2026.039.73
SFTDistillation Setting=Cross-Tokenizer KD2026.039.49
ULDDistillation Setting=Cross-Tokenizer KD2026.039.28
Student (GPT-2)Distillation Setting=Baseline2026.034.72