List your company

Question Answering on SQuAD KRE-curated version (EM/F1)

72.6F1 ScoreAdaCAD
4.06421.85739.6557.443Oct 14, 2025

Evaluation results

MethodLinks
AdaCADCategory=Decoding-Based, Backbone=Qwen3-8B2025.1072.667.4
ProbeRAGCategory=Training-Based, Backbone=Qwen3-8B2025.1071.555.7
COIECDCategory=Decoding-Based, Backbone=Qwen3-8B2025.1069.755.2
CANOECategory=Training-Based, Backbone=Qwen3-8B2025.1069.453.4
ParamMuteCategory=Training-Based, Backbone=Qwen3-8B2025.1069.252.1
ProbeRAGCategory=Training-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1068.453.3
ProbeRAGCategory=Training-Based, Backbone=Mistral-7B-v0.32025.1068.153.6
COIECDCategory=Decoding-Based, Backbone=LLaMA-3.1-8B-Instruct2025.106750.3
AdaCADCategory=Decoding-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1066.451.2
FaithfulRAGCategory=Prompt-Based, Backbone=Qwen3-8B2025.1066.250.2
Opin(Instr)Category=Prompt-Based, Backbone=LLaMA-3.1-8B-Instruct2025.106647.7
CANOECategory=Training-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1065.449.7
Full-ContextCategory=Baseline, Backbone=LLaMA-3.1-8B-Instruct2025.1064.546
Context-DPOCategory=Training-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1064.445.8
ParamMuteCategory=Training-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1064.248.5
Full-ContextCategory=Baseline, Backbone=Qwen3-8B2025.1063.844.9
Opin(Instr)Category=Prompt-Based, Backbone=Qwen3-8B2025.1063.846.1
Context-DPOCategory=Training-Based, Backbone=Qwen3-8B2025.1063.844.9
CADCategory=Decoding-Based, Backbone=Qwen3-8B2025.1063.644.5
AdaCADCategory=Decoding-Based, Backbone=Mistral-7B-v0.32025.1061.643.6
FaithfulRAGCategory=Prompt-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1061.350.2
CADCategory=Decoding-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1060.341.8
FaithfulRAGCategory=Prompt-Based, Backbone=Mistral-7B-v0.32025.1060.340.1
ParamMuteCategory=Training-Based, Backbone=Mistral-7B-v0.32025.1060.243.8
KRECategory=Prompt-Based, Backbone=LLaMA-3.1-8B-Instruct2025.1059.843.7
COIECDCategory=Decoding-Based, Backbone=Mistral-7B-v0.32025.1059.239.7
Opin(Instr)Category=Prompt-Based, Backbone=Mistral-7B-v0.32025.1058.137.4
CANOECategory=Training-Based, Backbone=Mistral-7B-v0.32025.1057.842.5
Context-DPOCategory=Training-Based, Backbone=Mistral-7B-v0.32025.1056.637.6
Full-ContextCategory=Baseline, Backbone=Mistral-7B-v0.32025.1056.437.5
KRECategory=Prompt-Based, Backbone=Mistral-7B-v0.32025.1052.633.9
CADCategory=Decoding-Based, Backbone=Mistral-7B-v0.32025.1051.432.1
KRECategory=Prompt-Based, Backbone=Qwen3-8B2025.1048.629.7
No-ContextCategory=Baseline, Backbone=LLaMA-3.1-8B-Instruct2025.108.91.2
No-ContextCategory=Baseline, Backbone=Mistral-7B-v0.32025.108.11
No-ContextCategory=Baseline, Backbone=Qwen3-8B2025.106.70.4