List your company
ResearchFormerly WizWand

Conversational response generation on INSCIT

33.2F1UniConv
12.19217.64623.128.554Oct 15, 2025
Updated 6mo ago

Evaluation results

MethodLinks
UniConvRAG=RAG, LLM=Mistral 7b, Training Strategy=Supervised Fine-tuning or RL Training2025.1033.2—
ChatR1-3bRAG=RAG, LLM=Qwen-3b, Training Strategy=Supervised Fine-tuning or RL Training2025.1033.285.5
ChatR1-7bRAG=RAG, LLM=Qwen-7b, Training Strategy=Supervised Fine-tuning or RL Training2025.1032.885.5
ChatR1 (w/o Rint.)RAG=RAG, LLM=Qwen-3b, Training Strategy=Supervised Fine-tuning or RL Training2025.1031.384.4
ChatRetriever +Mis.RAG=RAG, LLM=Mistral 7b, Training Strategy=Supervised Fine-tuning or RL Training2025.1030.3—
QR Search R1RAG=RAG, LLM=Qwen-3b, Training Strategy=Supervised Fine-tuning or RL Training2025.1027.584
Claude (DI)RAG=No, LLM=Claude, Training Strategy=Zero-Shot2025.1027—
conv-ANCE +Mis.RAG=RAG, LLM=Mistral 7b, Training Strategy=Supervised Fine-tuning or RL Training2025.1024.8—
CoT R1RAG=No, LLM=Qwen-3b, Training Strategy=Supervised Fine-tuning or RL Training2025.1024.184
ChatGPT (DI)RAG=No, LLM=GPT-3.5, Training Strategy=Zero-Shot2025.1022.881.1
IRCoTRAG=RAG, LLM=Qwen-3b, Training Strategy=Zero-Shot2025.1020.467.3
Qwen-Instr. (DI)RAG=No, LLM=Qwen-3b, Training Strategy=Zero-Shot2025.1017.958.1
SFTRAG=No, LLM=Qwen-3b, Training Strategy=Supervised Fine-tuning or RL Training2025.1016.956.9
Qwen-Instr. (CoT)RAG=No, LLM=Qwen-3b, Training Strategy=Zero-Shot2025.1016.462.4
Qwen-Instr. (RAG)RAG=RAG, LLM=Qwen-3b, Training Strategy=Zero-Shot2025.101349.3