List your company

Email Writing on Email Writing Weak User

0.3047Mean Total Edit DistanceDPO
0.2971960.3478480.39850.449152Jan 27, 2026

Evaluation results

MethodLinks
DPOTest-time User Model=Llama-3.3-70B-Instruct, Online learning phase=true2026.010.3047
DPOUser LLM=Qwen-3 32B, Seeds=3, Learning Phase=Online2026.010.3368
LateEnsembleTest-time User Model=Llama-3.3-70B-Instruct, Online learning phase=true2026.010.3428
LateEnsembleUser LLM=Qwen-3 32B, Seeds=3, Learning Phase=Online2026.010.3739
SFTTest-time User Model=Llama-3.3-70B-Instruct, Online learning phase=true2026.010.3866
EarlyEnsembleTest-time User Model=Llama-3.3-70B-Instruct, Online learning phase=true2026.010.4001
BaseTest-time User Model=Llama-3.3-70B-Instruct, Online learning phase=true2026.010.4239
SFTUser LLM=Qwen-3 32B, Seeds=3, Learning Phase=Online2026.010.4539
EarlyEnsembleUser LLM=Qwen-3 32B, Seeds=3, Learning Phase=Online2026.010.4864
BaseUser LLM=Qwen-3 32B, Seeds=3, Learning Phase=Online2026.010.4923