Uncovering Sentiment Analysis Circuit in Large Language Model

Large Language Models (LLMs) can perform sentiment analysis via natural language instructions, yet their predictions are highly sensitive to prompt phrasing. Prior work has shown that sentiment is encoded linearly in LLM representations, but the model's ability to utilize this information remains surprisingly fragile to prompt variations. To understand this behavior, we leverage Sparse Autoencoders (SAEs) to extract interpretable features from LLM activations and apply circuit-level analysis to uncover causal mechanisms underlying sentiment prediction. We identify a sentiment analysis circuit and find that prompt sensitivity may stem from task activation failure. Based on this insight, we propose a simple inference-time intervention method that amplifies circuit features to compensate for insufficient activation. Experiments across diverse datasets, templates, and languages show consistent improvements, offering an interpretable and training-free alternative to manual prompt engineering.