Domain Anchorage in GPT-4: A Computational Linguistic Analysis of Lexicographic Profiling and Its Implications for Unintended Information Dissemination

Simple item page
dc.contributor.authorChallappa, Lekha
dc.contributor.authorZhang, Jenevieve
dc.contributor.authorGarg, Rajiv
dc.date.accessioned2024-12-26T21:10:56Z
dc.date.available2024-12-26T21:10:56Z
dc.date.issued2025-01-07
dc.description.abstractOur study expands upon recent work explaining in-context learning as implicit Bayesian inference, where language models infer shared latent concepts from examples. We analyze GPT-4's semantic attention post-domain priming, using computational linguistics to quantify response similarity to lexicographically independent queries with the same intent. We assess potential privacy breaches from inadvertent domain anchorage, examining how attention and embedding layers process linguistic patterns. We hypothesize that domain-specific words receiving higher gradient updates can introduce bias, create semantic echo chambers, and oversimplify relationships. Grounded in Mohamed Zakaria Kurdi's frameworks, this research uses lexical, semantic, syntactic, and positional similarities to analyze GPT-4's vector transformations and attention distributions. By simulating domain-specific interactions through declarative primes and interrogative inputs, we highlight significant privacy and ethical concerns, as the model may share information across users due to domain anchorage.
dc.format.extent10
dc.identifier.doi10.24251/HICSS.2025.841
dc.identifier.isbn978-0-9981331-8-8
dc.identifier.other50cbd0b7-2ee7-4d8a-9c9d-8126304993e3
dc.identifier.urihttps://hdl.handle.net/10125/109692
dc.relation.ispartofProceedings of the 58th Hawaii International Conference on System Sciences
dc.rightsAttribution-NonCommercial-NoDerivatives 4.0 International
dc.rights.urihttps://creativecommons.org/licenses/by-nc-nd/4.0/
dc.subjectArtifical Intelligence Security: Ensuring Safety, Trustworthiness, and Responsibility in AI Systems
dc.subjectdomain anchorage, implicit profiling in ai, information dissemination, lexicographic similarity., semantic attention
dc.titleDomain Anchorage in GPT-4: A Computational Linguistic Analysis of Lexicographic Profiling and Its Implications for Unintended Information Dissemination
dc.typeConference Paper
dc.type.dcmiText
prism.startingpage7036

Files

Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
0686.pdf
Size:
269.98 KB
Format:
Adobe Portable Document Format
Download

Collections

Artifical Intelligence Security: Ensuring Safety, Trustworthiness, and Responsibility in AI Systems