SALSA: Single-pass Autoregressive LLM Structured Classification
This addresses a specific problem for users of LLMs in text classification applications, offering an incremental improvement over existing methods.
The paper tackles the underperformance of instruction-tuned Large Language Models on text classification benchmarks by introducing SALSA, a pipeline that combines structured prompting, class-to-token mapping, and parameter-efficient fine-tuning, achieving state-of-the-art results across diverse benchmarks.
Despite their impressive generalization capabilities, instruction-tuned Large Language Models often underperform on text classification benchmarks. We introduce SALSA, a coherent pipeline that combines structured prompting, class-to-token mapping, and parameter-efficient fine-tuning, thereby avoiding cold-start training. Each class label is mapped to a distinct output token, and prompts are constructed to elicit a single-token response. During inference, the model's output is projected only onto the logits of the relevant class tokens, enabling efficient and accurate classification in a single forward pass. SALSA achieves state-of-the-art results across diverse benchmarks, demonstrating its robustness and scalability for LLM-based classification applications.