paper-with-me

홈 › Papers

Identifying the Source of Generation for Large Language Models

2024-07-05 · Bumjin Park, Jaesik Choi

Large language models (LLMs) memorize text from several sources of documents. In pretraining, LLM trains to maximize the likelihood of text but neither receives the source of the text nor memorizes the source. Accordingly, LLM can not provide document information on the generated content, and users do not obtain any hint of reliability, which is crucial for factuality or privacy infringement. This work introduces token-level source identification in the decoding step, which maps the token representation to the reference document. We propose a bi-gram source identifier, a multi-layer perceptron with two successive token representations as input for better generalization. We conduct extensive experiments on Wikipedia and PG19 datasets with several LLMs, layer locations, and identifier sizes. The overall results show a possibility of token-level source identifiers for tracing the document, a crucial problem for the safe use of LLMs.

📄 PDF Abstract BibTeX arXiv:2407.12846

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

HINT An unsupervised approach for identifying Hierarchical Information Threads by analysing the network of related articles in a collection. In particular, HINT leverages article…

Similar Papers 제목 키워드 기반

Experiences in Resource Generation for Machine Translation through Crowdsourcing

2012-05-01 · LREC 2012 5 · Anoop Kunchukuttan, Shourya Roy, Pratik Patel, Kushal Ladha 외

The logistics of collecting resources for Machine Translation (MT) has always been a cause of concern for some of the resource deprived languages of the world. The recent advent of crowdsourcing platforms provides an opp…

Machine TranslationTranslation

LLM-Match: An Open-Sourced Patient Matching Model Based on Large Language Models and Retrieval-Augmented Generation

2025-03-17 · Xiaodi Li, Shaika Chowdhury, Chung Il Wi, Maria Vassilaki 외

Patient matching is the process of linking patients to appropriate clinical trials by accurately identifying and matching their medical records with trial eligibility criteria. We propose LLM-Match, a novel framework for…

RAGRetrieval-augmented Generation

Poisoning Programs by Un-Repairing Code: Security Concerns of AI-generated Code

2024-03-11 · Cristina Improta

AI-based code generators have gained a fundamental role in assisting developers in writing software starting from natural language (NL). However, since these large language models are trained on massive volumes of data c…

Code GenerationData PoisoningPosition

Multilingual Source Tracing of Speech Deepfakes: A First Benchmark

2025-08-06 · Xi Xuan, Yang Xiao, Rohan Kumar Das, Tomi Kinnunen arxiv

Recent progress in generative AI has made it increasingly easy to create natural-sounding deepfake speech from just a few seconds of audio. While these tools support helpful applications, they also raise serious concerns…

Source Attribution for Large Language Model-Generated Data

2023-10-01 · Jingtan Wang, Xinyang Lu, Zitong Zhao, Zhongxiang Dai 외

The impressive performances of Large Language Models (LLMs) and their immense potential for commercialization have given rise to serious concerns over the Intellectual Property (IP) of their training data. In particular,…

Language ModelingLanguage ModellingLarge Language Modelmodel