Article

Poisoning scientific knowledge using large language models

2023-11-08

Abstract excerpt

Biomedical knowledge graphs constructed from scientific literature have been widely used to validate biological discoveries and generate new hypotheses. Recently, large language models (LLMs) have demonstrated a strong ability to generate human-like text data. While most of these text data have been useful, LLM might also be used to generate malicious content. Here, we investigate whether it is possible that a mal...

Topics

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

Identifiers and source

Literature Corpus work: 7ffebeee-2939-533b-9edf-8839f4d51467
DOI: 10.1101/2023.11.06.565928

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Poisoning scientific knowledge using large language modelsDOI 10.1101/2023.11.06.565928Read

Detection of patient metadata in published articles for genomic epidemiology using machine learning and large language modelsSemantic proximity Read

Simulate Scientific Reasoning with Multiple Large Language Models: An Application to Alzheimer’s Disease Combinatorial TherapySemantic proximity Read

SKiM-GPT: Combining Biomedical Literature-Based Discovery with Large Language Model Hypothesis EvaluationSemantic proximity Read

LORE: A Literature Semantics Framework for Evidenced Disease-Gene Pathogenicity Prediction at ScaleSemantic proximity Read

Faithful AI in Medicine: A Systematic Review with Large Language Models and BeyondSemantic proximity Read

Dual Inference Routes for Natural Language Hypothesis Generation in Multimodal Drug Discovery: An Early Experimentation StudySemantic proximity Read

Large language models are universal biomedical simulatorsSemantic proximity Read

Comparing scientific abstracts generated by ChatGPT to original abstracts using an artificial intelligence output detector, plagiarism detector, and blinded human reviewersSemantic proximity Read

A Biosecurity Agent for Lifecycle LLM Biosecurity AlignmentSemantic proximity Read

Can large language models reliably extract human disease genes from full-text scientific literature?Semantic proximity Read

A dataset for evaluating clinical research claims in large language modelsSemantic proximity Read

Zero-shot biological reasoning with open-weights large language models reproduces CRISPR screen based prediction of synthetic lethal interactionsSemantic proximity Read