List your company

Counterfactual Generation on SNLI Premise

0.759LFRFIZLE
0.110040.278520.4470.61548Jan 4, 2026

Evaluation results

MethodLinks
FIZLEExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B2026.010.7590.90841.54
Causal What-IfsExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B2026.010.750.86441.07
iFlip-NLExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B, Feedback Type=Natural Language2026.010.730.78977
iFlip-SHAPExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B, Feedback Type=Feature Attribution2026.010.680.757.99
iFlip-ConfExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B, Feedback Type=Confidence2026.010.6650.68862.94
iFlipBackbone LLM=OLMo2-7B, Feedback Type=Confidence2026.010.660.85453.72
iFlipBackbone LLM=OLMo2-7B, Feedback Type=Natural Language2026.010.6380.879148.41
CGGExplained Model=RoBERTa-base, Generator Model=LLaMA 3.3-70B2026.010.6350.89359.23
iFlipBackbone LLM=OLMo2-7B, Feedback Type=Grad×Input2026.010.6240.85456.02
iFlipBackbone LLM=OLMo2-7B, Feedback Type=AttnLRP2026.010.5960.85155.3
iFlipBackbone LLM=Qwen3-32B, Feedback Type=Natural Language2026.010.5960.89568.16
iFlipBackbone LLM=OLMo2-7B, Feedback Type=SHAP2026.010.5870.84756.81
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=Confidence2026.010.580.76358.65
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=Grad×Input2026.010.580.75155.38
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=Natural Language2026.010.5660.82169.45
iFlipBackbone LLM=OLMo2-7B, Feedback Type=LIME2026.010.5580.86658.18
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=LIME2026.010.5580.75957.87
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=AttnLRP2026.010.5280.74356.42
iFlipBackbone LLM=LLaMA3.3-70B, Feedback Type=SHAP2026.010.5220.74357.62
BAEExplained Model=RoBERTa-base2026.010.5180.894103.32
Causal What-IfsBackbone LLM=Qwen3-32B2026.010.4820.84236.14
iFlipBackbone LLM=Qwen3-32B, Feedback Type=Confidence2026.010.4660.89562.69
iFlipBackbone LLM=Qwen3-32B, Feedback Type=LIME2026.010.4580.89162.98
FIZLEBackbone LLM=Qwen3-32B2026.010.4540.90138.66
iFlipBackbone LLM=Qwen3-32B, Feedback Type=SHAP2026.010.4520.87858.6
iFlipBackbone LLM=Qwen3-32B, Feedback Type=AttnLRP2026.010.4480.88258.41
iFlipBackbone LLM=Qwen3-32B, Feedback Type=Grad×Input2026.010.440.88158.66
Causal What-IfsBackbone LLM=OLMo2-7B2026.010.4130.85447.27
FIZLEBackbone LLM=LLaMA3.3-70B2026.010.4020.90341.54
Causal What-IfsBackbone LLM=LLaMA3.3-70B2026.010.3720.81638.97
CGGBackbone LLM=LLaMA3.3-70B2026.010.3070.84651.62
FIZLEBackbone LLM=OLMo2-7B2026.010.3050.73173.94
Polyjuice2026.010.2810.77296.95
PolyjuiceExplained Model=RoBERTa-base2026.010.2120.77296.95
CGGBackbone LLM=Qwen3-32B2026.010.20.88150.31
CGGBackbone LLM=OLMo2-7B2026.010.1690.88257.03
BAE2026.010.1350.96682.98