List your company
ResearchFormerly WizWand

Veracity Prediction on LIAR RAW

50.59Macro F1S-EGS
13.72223.293532.86542.4365May 21, 2025Jul 13, 2025Sep 5, 2025Oct 29, 2025Dec 21, 2025Feb 13, 2026Apr 8, 2026
Updated 6mo ago

Evaluation results

MethodLinks
S-EGSx -> y=c → v; exp, Model (Backbone)=LLaMA2-7B, No Closed-Source API Dependency=true, # Distill Explanations for RAW-FC=4652025.1150.5952.4550.39
SFTx -> y=c → v; exp, Model (Backbone)=LLaMA2-7B, No Closed-Source API Dependency=true, # Distill Explanations for RAW-FC=02025.1143.0548.3846.83
RAFTSReproduced=false2025.05424737
HiSS (Google)x -> y=c → v; exp, Model (Backbone)=ChatGPT, No Closed-Source API Dependency=false, utilizing search API=Google2025.1137.546.831.3
HiSSReproduced=false2025.05374631
G-DefenseLLaMA3.1Approach category=Ours2026.0432.4934.1732.37
G-DefenseApproach category=Ours2026.0431.5533.0931.55
L-DenfenseLLaMA2Approach category=LLM-based approach2026.0431.431.6331.71
L-Defense (Llama2)Backbone=Llama2, Reproduced=false2025.05313131
G-DefensebackgroundApproach category=Ours2026.0430.8931.531.28
L-DenfenseGPT-3.5Approach category=LLM-based approach2026.0430.5330.5532.2
L-Denfensex -> y=c; evi → v; exp, Model (Backbone)=ChatGPT + Roberta-Large, No Closed-Source API Dependency=false, # Distill Explanations for RAW-FC=32,2402025.1130.5330.5532.2
FactLLaMAknowApproach category=LLM-based approach2026.0430.4432.4632.05
FactLLaMA (Google)x -> y=c; evi → v, Model (Backbone)=LLaMA2-7B, No Closed-Source API Dependency=false, utilizing search API=Google2025.1130.4432.4632.05
G-Defensedecomp+Approach category=Ours2026.0430.3632.0330.23
FactLLaMAReproduced=false2025.05303232
L-Defense (ChatGPT)Backbone=ChatGPT, Reproduced=false2025.05303032
FactLLaMAApproach category=LLM-based approach2026.0429.9832.3231.57
G-DefensehyperApproach category=Ours2026.0429.7531.2129.53
L-Defense (Reproduced)Reproduced=true2025.05292929
CofCEDApproach category=Traditional approach2026.0428.9329.4829.55
MUSERApproach category=Traditional approach2026.0426.5227.826.58
GenFEApproach category=Traditional approach2026.0426.4928.0126.16
RAVx -> y=c; evi → v; exp, Model (Backbone)=LLaMA-3.1-70B-Instruct x3, No Closed-Source API Dependency=true2025.1125.4——
GPT-3.5claimApproach category=LLM-based approach2026.0425.1125.4127.33
ChatGPTx -> y=c → v; exp, No Closed-Source API Dependency=false2025.1125.1125.4127.33
SBERT-FCApproach category=Traditional approach2026.0422.1924.0922.07
GPT-3.5fullApproach category=LLM-based approach2026.0421.929.6423.57
ChatGPTx -> y=c, evi → v; exp, No Closed-Source API Dependency=false2025.1121.929.6423.57
dEFENDApproach category=Traditional approach2026.0417.5123.0918.56
GenFE-MTApproach category=Traditional approach2026.0415.1518.5519.9
LLaMA2claimApproach category=LLM-based approach2026.0415.1417.1117.37