List your company

Counterfactual Generation on AG News

0.915LFRiFlip-NL
0.117320.324410.53150.73859Jan 4, 2026

Evaluation results

MethodLinks
iFlip-NLExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B, Feedback Type=Natural Language2026.010.9150.58357.66
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=Natural Language2026.010.90.52763.86
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=Confidence2026.010.890.46844.38
iFlip-SHAPExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B, Feedback Type=Feature Attribution2026.010.8850.54257.69
iFlip-ConfExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B, Feedback Type=Confidence2026.010.880.57555.55
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=SHAP2026.010.860.48146.07
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=AttnLRP2026.010.8580.47745.78
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=Grad×Input2026.010.8520.47946.12
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=LIME2026.010.8520.47544.2
iFlipBackbone LLM=OLMo2-7B, Feedback Type=SHAP2026.010.8020.48842.95
iFlipBackbone LLM=OLMo2-7B, Feedback Type=Confidence2026.010.7840.50341.41
iFlipBackbone LLM=OLMo2-7B, Feedback Type=Grad×Input2026.010.7730.49850.32
iFlipBackbone LLM=OLMo2-7B, Feedback Type=AttnLRP2026.010.7680.546.61
CGGExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B2026.010.7480.542130.31
iFlipBackbone LLM=Qwen3-32B, Feedback Type=AttnLRP2026.010.7460.51344.15
iFlipBackbone LLM=Qwen3-32B, Feedback Type=LIME2026.010.7420.52744.52
iFlipBackbone LLM=Qwen3-32B, Feedback Type=SHAP2026.010.7360.51944.02
iFlipBackbone LLM=OLMo2-7B, Feedback Type=Natural Language2026.010.7350.51544.73
iFlipBackbone LLM=Qwen3-32B, Feedback Type=Natural Language2026.010.7320.53649.68
iFlipBackbone LLM=OLMo2-7B, Feedback Type=LIME2026.010.730.50444.32
iFlipBackbone LLM=Qwen3-32B, Feedback Type=Grad×Input2026.010.730.52244.33
iFlipBackbone LLM=Qwen3-32B, Feedback Type=Confidence2026.010.7260.52141.59
Causal What-IfsExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B2026.010.6560.44246.18
Causal What-IfsBackbone LLM=LLaMA3.3-70B2026.010.6360.46143.56
CGGBackbone LLM=LLaMA3.3-70B2026.010.5260.694122.4
CGGBackbone LLM=OLMo2-7B2026.010.510.38208.56
BAEExplained Model=RoBERTa-base2026.010.4430.813246.61
Causal What-IfsBackbone LLM=Qwen3-32B2026.010.4360.5753.5
FIZLEBackbone LLM=LLaMA3.3-70B2026.010.4160.70756.59
Causal What-IfsBackbone LLM=OLMo2-7B2026.010.40.60737.54
CGGBackbone LLM=Qwen3-32B2026.010.3640.69797.1
FIZLEExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B2026.010.3450.70756.59
FIZLEBackbone LLM=OLMo2-7B2026.010.3080.58477.38
BAE2026.010.2270.764295.96
PolyjuiceExplained Model=RoBERTa-base2026.010.1650.74493.22
FIZLEBackbone LLM=Qwen3-32B2026.010.1640.86461.46
Polyjuice2026.010.1480.74493.22