List your company

Multi-step Reasoning Question Answering on ConFiQA MR (test)

91.3F1 ScoreProbeRAG
1.44424.77248.171.428Oct 14, 2025

Evaluation results

MethodLinks
ProbeRAGCategory=Training-Based, Backbone=Qwen3-8B2025.1091.389
ProbeRAGCategory=Training-Based, Backbone=Mistral-7B-v0.32025.1090.888.2
ProbeRAGCategory=Training-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1089.787
ParamMuteCategory=Training-Based, Backbone=Mistral-7B-v0.32025.1086.482.1
CANOECategory=Training-Based, Backbone=Mistral-7B-v0.32025.1084.781.9
CANOECategory=Training-Based, Backbone=Qwen3-8B2025.1084.680.7
ParamMuteCategory=Training-Based, Backbone=Qwen3-8B2025.1083.279.4
CANOECategory=Training-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1080.272.6
Context-DPOCategory=Training-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1078.566.9
ParamMuteCategory=Training-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1073.261.2
AdaCADCategory=Decoding-Based, Backbone=Qwen3-8B2025.1072.564.3
COIECDCategory=Decoding-Based, Backbone=Qwen3-8B2025.1071.563.8
Opin(Instr)Category=Prompt-Based, Backbone=Qwen3-8B2025.1069.759.5
KRECategory=Prompt-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1068.758.9
KRECategory=Prompt-Based, Backbone=Qwen3-8B2025.1068.459
Full-ContextCategory=Baseline, Backbone=Qwen3-8B2025.1066.155.1
Opin(Instr)Category=Prompt-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1065.954
Context-DPOCategory=Training-Based, Backbone=Qwen3-8B2025.1065.955
CADCategory=Decoding-Based, Backbone=Qwen3-8B2025.1064.853.3
FaithfulRAGCategory=Prompt-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1062.453.1
FaithfulRAGCategory=Prompt-Based, Backbone=Qwen3-8B2025.1062.354.2
KRECategory=Prompt-Based, Backbone=Mistral-7B-v0.32025.1060.945.3
Opin(Instr)Category=Prompt-Based, Backbone=Mistral-7B-v0.32025.1057.852.5
FaithfulRAGCategory=Prompt-Based, Backbone=Mistral-7B-v0.32025.1056.542.1
COIECDCategory=Decoding-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1050.943.3
Context-DPOCategory=Training-Based, Backbone=Mistral-7B-v0.32025.1050.931.9
Full-ContextCategory=Baseline, Backbone=LLaMA-3.1-8B-Instruct2025.1050.341.3
AdaCADCategory=Decoding-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1046.742.5
CADCategory=Decoding-Based, Backbone=LLaMA-3.1-8B-Instruct2025.104031.3
AdaCADCategory=Decoding-Based, Backbone=Mistral-7B-v0.32025.1039.629.4
COIECDCategory=Decoding-Based, Backbone=Mistral-7B-v0.32025.1039.326.3
Full-ContextCategory=Baseline, Backbone=Mistral-7B-v0.32025.1037.821.5
CADCategory=Decoding-Based, Backbone=Mistral-7B-v0.32025.1027.512.8
No-ContextCategory=Baseline, Backbone=Qwen3-8B2025.1082.8
No-ContextCategory=Baseline, Backbone=LLaMA-3.1-8B-Instruct2025.106.11.9
No-ContextCategory=Baseline, Backbone=Mistral-7B-v0.32025.104.90.5