LabelFusion: Learning to Fuse LLMs and Transformer Classifiers for Robust Text Classification
This addresses the problem of improving accuracy and cost-efficiency in text classification for users needing robust multi-class and multi-label solutions, though it is incremental as it combines existing methods.
The paper tackles robust text classification by learning to fuse traditional transformer classifiers with Large Language Models (LLMs), achieving 92.4% accuracy on AG News and 92.3% on Reuters 21578 topic classification.
LabelFusion is a fusion ensemble for text classification that learns to combine a traditional transformer-based classifier (e.g., RoBERTa) with one or more Large Language Models (LLMs such as OpenAI GPT, Google Gemini, or DeepSeek) to deliver accurate and cost-aware predictions across multi-class and multi-label tasks. The package provides a simple high-level interface (AutoFusionClassifier) that trains the full pipeline end-to-end with minimal configuration, and a flexible API for advanced users. Under the hood, LabelFusion integrates vector signals from both sources by concatenating the ML backbone's embeddings with the LLM-derived per-class scores -- obtained through structured prompt-engineering strategies -- and feeds this joint representation into a compact multi-layer perceptron (FusionMLP) that produces the final prediction. This learned fusion approach captures complementary strengths of LLM reasoning and traditional transformer-based classifiers, yielding robust performance across domains -- achieving 92.4% accuracy on AG News and 92.3% on 10-class Reuters 21578 topic classification -- while enabling practical trade-offs between accuracy, latency, and cost.