List your company

Intent Inference on CSID (test)

80LLM Judge AgreementSCRIPTMIND-FT
14.4831.4948.565.51Jan 20, 2026

Evaluation results

MethodLinks
SCRIPTMIND-FTbackbone=EEVE-Korean-Instruct-10.8B, fine-tuned=true2026.0180
chatgpt-4o-latestzero-shot=true2026.0178
gemini-2.0-flashzero-shot=true2026.0177
SCRIPTMIND-FTbackbone=EXAONE-3.5-2.4B-Instruct, fine-tuned=true2026.0173
claude-3-5-haikuzero-shot=true2026.0173
EXAONE-3.0-7.8B-Instructzero-shot=true2026.0163
SCRIPTMIND-FTbackbone=Llama-3.2-1B-Instruct, fine-tuned=true2026.0157
Midm-2.0-Base-Instructzero-shot=true2026.0157
Llama-3.1-8B-Instructzero-shot=true2026.0154
SCRIPTMIND-FTbackbone=Llama-3.1-8B-Instruct, fine-tuned=true2026.0154
SCRIPTMIND-FTbackbone=Midm-2.0-Mini-Instruct, fine-tuned=true2026.0153
EEVE-Korean-Instruct-10.8Bzero-shot=true2026.0152
EXAONE-3.5-2.4B-Instructzero-shot=true2026.0151
SCRIPTMIND-FTbackbone=SOLAR-10.7B-Instruct, fine-tuned=true2026.0150
SCRIPTMIND-FTbackbone=Midm-2.0-Base-Instruct, fine-tuned=true2026.0150
SOLAR-10.7B-Instructzero-shot=true2026.0144
Midm-2.0-Mini-Instructzero-shot=true2026.0129
Llama-3.2-1B-Instructzero-shot=true2026.0117