AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

ZipLJP: Zipped Information Processor for Legal Judgment Prediction

Fanghao Lou, Qiqi Wang, Guanyu Chen, Kaiqi Zhao, Huijia Li

AAAI 2026 · Large language models › Language, safety and interpretability · approach: LLM unlearning and benchmarks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add ZipLJP compression to Qwen2.5-3B fine-tuningQwen2.5-3B · MultiLJP chargeAccuracy: 88.7 -> 90.1Improved
add ZipLJP compression to LLaMA3.2-3B fine-tuningLLaMA3.2-3B · CAIL2018 charge and law articleAccuracy: charge 77.4 -> 78.3; law article 74.3 -> 73.5Mixed
add ZipLJP compression to LLaMA3-8B fine-tuningLLaMA3-8B · CAIL2018 prison termAccuracy: 57.8 -> 55.1Hurt
add ZipLJP compression to LawyerLLaMA-13B-V2 fine-tuningLawyerLLaMA-13B-V2 · CAIL2018 chargeAccuracy: 82.4 -> 86.4Improved
use BGE embedding model instead of Legal-BERT for sentence selection (ablation)ZipLJP (Qwen2-7B) · MultiLJP law articleAccuracy: 79.7 -> 93.6Improved
use BGE embedding model instead of M3E (ablation)ZipLJP (Qwen2-7B) · MultiLJP law articleAccuracy: 93.3 -> 93.6Improved
use ZipLJP compression instead of BigBird sparse attention (ablation)Qwen2-7B · MultiLJP law articleAccuracy: 78.1 -> 93.6Improved
use ZipLJP compression instead of LLMLingua-2 (ablation)Qwen2-7B · MultiLJP law articleAccuracy: 83.1 -> 93.6Improved
use ZipLJP compression instead of Qwen-Plus abstractive summary (ablation)Qwen2-7B · MultiLJP prison termAccuracy: 57.3 -> 64.5Improved
vary retained proportion k of sentences (below vs above 0.2) (ablation)ZipLJP · MultiLJPAccuracy and F1: k < 0.2 low performance; k > 0.2 similar high performanceMixed
remove the most similar 20% or 60% of text before prediction (ablation)ZipLJP · MultiLJPAccuracy and F1Hurt
add ZipLJP to zero-shot DeepSeekDeepSeek-V3.2-Exp · MultiLJP law articleAccuracy and Macro-F: Acc. 89.1 -> 91.1; Ma-F 78.7 -> 77.3Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy