Mitigating Spurious Correlations in NLI via LLM-Synthesized Counterfactuals and Dynamic Balanced Sampling

Natural Language Inference (NLI) models frequently rely on spurious correlations rather than semantic reasoning. Existing mitigation strategies often incur high annotation costs or trigger catastrophic forgetting during fine-tuning. We propose an automated, scalable pipeline to address these limitations. First, we introduce Log-Frequency LMI (LF-LMI) to accurately detect semantic artifacts. Second, we generate a high-quality synthetic contrast set via an LLM-synthesis pipeline with multi-judge verification. Finally, we introduce Dynamic Balanced Sampling, a training strategy that rotates the original data distribution to prevent forgetting. Our method improves consistency on a challenging benchmark from 63.5% to 81.0% while maintaining 88.4% in-domain accuracy, significantly outperforming naive fine-tuning.