As large language models (LLMs) are increasingly deployed for text generation, watermarking has become essential for authorship attribution, intellectual property protection, and misuse detection. While existing watermarking methods perform well in high-resource languages, their robustness in low-resource languages remains underexplored. This work presents the first systematic evaluation of state-of-the-art text watermarking methods: KGW, Exponential Sampling (EXP), and Waterfall, for Bangla LLM text generation under cross-lingual round-trip translation (RTT) attacks. Under benign conditions, KGW and EXP achieve high detection accuracy (>88%) with negligible perplexity and ROUGE degradation. However, RTT causes detection accuracy to collapse below RTT causes detection accuracy to collapse to 9-13%, indicating a fundamental failure of token-level watermarking. To address this, we propose a layered watermarking strategy that combines embedding-time and post-generation watermarks. Experimental results show that layered watermarking improves post-RTT detection accuracy by 25-35%, achieving 40-50% accuracy, representing a 3$\times$ to 4$\times$ relative improvement over single-layer methods, at the cost of controlled semantic degradation. Our findings quantify the robustness-quality trade-off in multilingual watermarking and establish layered watermarking as a practical, training-free solution for low-resource languages such as Bangla. Our code and data will be made public.
A Survey of Text Watermarking in the Era of Large Language Models
Tsinghua University（清华大学）；The Chinese University of Hong Kong（香港中文大学）；The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））；Beijing University of Posts and Telecommunications（北京邮电大学）；University of Illinois Chicago（芝加哥大学伊利诺伊分校）
Beyond Clean Text: Evaluating Encoder and Decoder Robustness for Bangla Event Detection in Noisy Text
Jahangirnagar University（贾汉吉尔纳加尔大学）；Rajshahi University of Engineering & Technology（拉杰沙希工程技术大学）；Bangladesh University of Engineering and Technology（孟加拉国工程技术大学）
Bangladesh University of Engineering Technology (BUET)（孟加拉国工程技术大学）；AWS AI Labs（亚马逊云科技人工智能实验室）；University of Rochester（罗切斯特大学）
In this work, we introduce BanglaBERT, a BERT-based Natural Language Understanding (NLU) model pretrained in Bangla, a widely spoken yet low-resource language in the NLP literature. To pretrain BanglaBERT, we collect 27.5 GB of Bangla pretraining data (dubbed `Bangla2B+') by crawling 110 popular Bangla sites. We introduce two downstream task datasets on natural language inference and question answering and benchmark on four diverse NLU tasks covering text classification, sequence labeling, and span prediction. In the process, we bring them under the first-ever Bangla Language Understanding Ben…
Revealing Weaknesses in Text Watermarking Through Self-Information Rewrite Attacks
University of British Columbia（不列颠哥伦比亚大学）；Vector Institute for AI（向量人工智能研究所）；University of Cambridge（剑桥大学）；Canada CIFAR AI Chair（加拿大CIFAR人工智能主席）；NSERC CRC Chair（NSERC加拿大研究主席）
Towards Codable Watermarking for Injecting Multi-bits Information to LLMs
National Key Laboratory for Multimedia Information Processing, School of Computer Science, Peking University（北京大学计算机学院多媒体信息处理国家重点实验室）；Gaoling School of Artificial Intelligence, Renmin University of China（中国人民大学高瓴人工智能学院）；Pattern Recognition Center, WeChat AI, Tencent Inc.（腾讯公司微信AI模式识别中心）；DeepSeek AI（深度求索人工智能公司）
The Bangla language is the seventh most spoken language, with 265 million native and non-native speakers worldwide. However, English is the predominant language for online resources and technical knowledge, journals, and documentation. Consequently, many Bangla-speaking people, who have limited command of English, face hurdles to utilize English resources. To bridge the gap between limited support and increasing demand, researchers conducted many experiments and developed valuable tools and techniques to create and process Bangla language materials. Many efforts are also ongoing to make it eas…
Overview of PAN 2026: Voight-Kampff Generative AI Detection, Text Watermarking, Multi-Author Writing Style Analysis, Generative Plagiarism Detection, and Reasoning Trajectory Detection
Friedrich Schiller University Jena（耶纳弗里德里希·施勒格尔大学）；University of Applied Sciences BFI（应用科学大学BFI）；Mohamed bin Zayed University of Artificial Intelligence（马尔代夫 bin Zayed 人工智能大学）；University of Kassel（卡塞尔大学）；INSAIT, Sofia University ``St. Kliment Ohridski''（INSAIT，索菲亚大学『圣克莱门特·奥赫里迪斯』）；University of Innsbruck（因斯布鲁克大学）
PAN 2026旨在通过客观评估推进计算风格学和文本取证，涵盖生成式AI检测、文本水印、多作者写作分析、抄袭检测及推理轨迹检测等五个新旧任务。
Institute of Information Engineering, Chinese Academy of Sciences（中国科学院信息工程研究所）；School of Cyber Security, University of Chinese Academy of Sciences（中国科学院大学网络与信息安全学院）
Block-wise Codeword Embedding for Reliable Multi-bit Text Watermarking
Department of AI, DGIST, Daegu, Republic of Korea（韩国大邱科学技术院人工智能系）；Department of EECS, DGIST, Daegu, Republic of Korea（韩国大邱科学技术院电子工程与计算机科学系）；ICT convergence, University of Ulsan, Ulsan, Republic of Korea（韩国釜山大学信息通信融合学院）
From Trade-off to Synergy: A Versatile Symbiotic Watermarking Framework for Large Language Models
Institute of Information Engineering, Chinese Academy of Sciences（中国科学院信息工程研究所）；School of Cyber Security, University of Chinese Academy of Sciences（中国科学院大学网络安全学院）；Hainan Province Fang Binxing Academician Workstation（海南省方斌星院士工作站）
Breaking the Silence: A Dataset and Benchmark for Bangla Text-to-Gloss Translation
IIT, University of Dhaka, Bangladesh（达卡大学理工学院，孟加拉国）；University of Maryland, Baltimore County, USA（马里兰大学巴尔的摩县分校，美国）；SUNY, Old Westbury, USA（SUNY，美国奥尔德韦斯特伯里）
Transformer-based Text Classification on Unified Bangla Multi-class Emotion Corpus
School of Management Science and Engineering, Shandong University of Finance and Economics（山东财经大学管理科学与工程学院）；College of Computer Science and Software Engineering, Shenzhen University（深圳大学计算机与软件学院）
Universal Networking Language (UNL) is a declarative formal language that is used to represent semantic data extracted from natural language texts. This paper presents a novel approach to converting Bangla natural language text into UNL using a method known as Predicate Preserving Parser (PPP) technique. PPP performs morphological, syntactic and semantic, and lexical analysis of text synchronously. This analysis produces a semantic-net like structure represented using UNL. We demonstrate how Bangla texts are analyzed following the PPP technique to produce UNL documents which can then be transl…
