Get Started
Home
Topics
Search
Library
Research questionHow can legal judgment prediction models avoid guilty outcomes when case narratives favor the prosecution?Legal judgment prediction systems often learn from documents written from a prosecutorial perspective and datasets dominated by guilty labels. These conditions can make models treat the prosecution’s narrative as objective and overlook plausible innocence.
AI
Alignment & Safety
Multi-agent Systems
Natural Language Processing
Reasoning
Latest papersRecent research connected to this question, newest first.OBJECTION! Lawyer Agents Mitigate Guilty Bias in Legal Judgment PredictionThe evidence concerns an inference-time adversarial lawyer-agent pipeline evaluated on a 3.4k-case real-world Natural Innocent dataset. The reported findings are limited to the evaluated legal judgment prediction setting and its false-guilty outcomes.research paper · Sep 2, 2026
Related questions
How can we detect internally incoherent language-model forecasts before relying on them for consequential decisions?How can evaluators distinguish missing knowledge from miscalibrated outputs in language models?How can we evaluate LLM reasoning quality beyond final-answer accuracy across deployment contexts?How can LLM prompts be automatically refined from recurring reasoning errors without laborious manual engineering?