paper

Towards Explainable Adjudicative Variance: Quantifying Judicial Discretion via Gated Multi-Task Learning

paperactiveprovisionaltowards-explainable-adjudicative-variance-quantifying-judicial-discretion-via-gated-multi-task-learning-3b508513·1 events·first seen 3d ago

Aliases: Towards Explainable Adjudicative Variance: Quantifying Judicial Discretion via Gated Multi-Task Learning

Co-occurring entities

Gemma-4 E4B-it LoRA

More like this (12)

Can LLMs Judge Better Than They Generate? Evaluating Task Asymmetry, Mechanistic Interpretability and Transferability for In-Context QA Sparse Subspace-to-Expert Sharing for Task-Agnostic Continual Learning Multi-Task Learning Multi-Task Bayesian In-Context Learning Cognitive Episodes in LLM Reasoning Traces Enable Interpretable Human Item Difficulty Prediction Listening with Attention: Entropy-Guided Explainability for Transformer-Based Audio Models Generative Explainability for Next-Generation Networks: LLM-Augmented XAI with Mutual Feature Interactions Generative Explainability for Next-Generation Networks: LLM-Augmented XAI with Mutual Feature Interactions Hierarchical Advantage Weighting for Online RL Fine-Tuning of VLAs from Sparse Episode Outcomes Same Evidence, Different Answer: Auditing Order Sensitivity in Multimodal Large Language Models Exploring Adversarial Robustness and Safety Alignment in Multilingual Multi-Modal Large Language Models Uncertainty Quantification for Computer-Use Agents: A Benchmark across Vision-Language Models and GUI Grounding Datasets

Recent events (1)

4arXiv · cs.CL·3d ago·source ↗

Judge-Aware Gated Multi-Task Learning achieves state-of-the-art on UK Employment Tribunal outcome prediction

Researchers propose a Judge-Aware Gated Multi-Task Learning architecture for legal outcome prediction that explicitly disentangles factual case merits from judicial discretion via a gated fusion mechanism conditioned on judge identity. Evaluated on 13,937 UK Employment Tribunal decisions, the approach outperforms supervised fine-tuning of a Gemma-4 26B backbone while requiring an order of magnitude fewer trainable parameters. The key finding is that differentiable structured composition of identity signals outperforms prompt-based composition over a much larger generative model, suggesting conditioning interface choice dominates scale for identity-conditioned classification tasks.

Evaluation and Benchmarking Gemma-4 E4B-it Towards Explainable Adjudicative Variance: Quantifying Judicial Discretion via Gated Multi-Task Learning LoRA