Article

Adversarial BERT with Robustness-Aware Decoding for Noise-Resilient Legal Document Classification and Summarization

2026-08-04

Abstract excerpt

Operational legal pipelines ingest scanned filings, heterogeneous citation conventions, and documents whose boilerplate has been rewritten, yet the encoders behind them are evaluated almost entirely on clean, born-digital benchmarks. We take the position that noise in legal text is structured rather than random, and we build a framework on that premise. LegalStruct-Attack defines a perturbation space over five fam...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: 17674eda-a873-5032-82a4-9d10492b7b07
DOI: 10.20944/preprints202608.0177.v1

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Adversarial BERT with Robustness-Aware Decoding for Noise-Resilient Legal Document Classification and SummarizationDOI 10.20944/preprints202608.0177.v1Read

Adversarial Text Generation with Dynamic Contextual PerturbationSemantic proximity Read

The Cognitive Fingerprint Problem: Transformer-LSTM Perplexity Geometry for Fair and Adversarially Robust AI Text DetectionSemantic proximity Read

LLM-as-Critic: Contrastive and Adversarial Strategies for Authentic Text VerificationSemantic proximity Read

Understanding Transformer-Based OCR for Medieval Manuscripts: A Systematic Ablation Study and Inspection AnalysisSemantic proximity Read

Cross-LLM Paraphrase Laundering: A Register-Controlled Evaluation of Fake-News DetectorsSemantic proximity Read

Mixture of Detectors: A Compact View of Machine-Generated Text DetectionSemantic proximity Read

SAS Multimetric Tribunal: Experimental Validation of a Three-Zone Structural Classification System for Hallucination Detection in AI-Generated TextSemantic proximity Read

Normalization Is Necessary but Insufficient: A Cross-Class Audit of Glyph-Level Obfuscation and Attributive Reframing Against LLM Content-Safety GuardsSemantic proximity Read

Vernacular Remote Control: Adversarial Robustness Testing of Six Commercial AI Text Detectors on Provenance-Guaranteed Pre-AI ProseSemantic proximity Read

Responsible AI in NLP: GUS-Net Span-Level Bias Detection Dataset and Benchmark for Generalizations, Unfairness, and StereotypesSemantic proximity Read

Cross-Dataset Generalization in Urdu Fake News Detection: An Empirical Study with XLM-RoBERTa and a Length Confound AnalysisSemantic proximity Read

Exploring Robustness of Multilingual LLMs on Real-World Noisy DataSemantic proximity Read