A gated multi-task architecture with outcome taxonomy achieves state-of-the-art legal outcome prediction on 13,937 UK Employment Tribunal cases by disentangling judge identity, using far fewer parameters than generative fine-tuning of a 26B model.
findings-eacl.44/
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Towards Explainable Adjudicative Variance: Quantifying Judicial Discretion via Gated Multi-Task Learning
A gated multi-task architecture with outcome taxonomy achieves state-of-the-art legal outcome prediction on 13,937 UK Employment Tribunal cases by disentangling judge identity, using far fewer parameters than generative fine-tuning of a 26B model.