SOURCE-LINKED INTELLIGENCE

From Two Passes to One: Compact and Efficient Target-Stance Extraction

arXiv · AI, language, vision and robotics · article · Sep 5, 2026 · UTC

Target-Stance Extraction (TSE) is the task of predicting both the target (or topic) of an author's writing and the author's stance toward it. Existing approaches to TSE use a sequential pipeline of two separate neural models: one to identify the target and another to determine the stance. We present a one-pass, joint architecture that predicts both in a single forward pass, reducing trainable parameters by nearly 50% with only a 4-7 F1 point tradeoff in performance. We further demonstrate that standard target-scrubbing practices artificially suppress target prediction accuracy. Retaining expli

Read original source ↗ Open in workspace

recordType

paper

region

Global