Domain Anchorage in GPT-4: A Computational Linguistic Analysis of Lexicographic Profiling and Its Implications for Unintended Information Dissemination
Location
Hilton Waikoloa Village, Hawaii
Event Website
https://hicss.hawaii.edu/
Start Date
7-1-2025 12:00 AM
End Date
10-1-2025 12:00 AM
Description
Our 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.
Recommended Citation
Challappa, Lekha; Zhang, Jenevieve; and Garg, Rajiv, "Domain Anchorage in GPT-4: A Computational Linguistic Analysis of Lexicographic Profiling and Its Implications for Unintended Information Dissemination" (2025). Hawaii International Conference on System Sciences 2025 (HICSS-58). 5.
https://aisel.aisnet.org/hicss-58/st/safety_trustworthy_and_responsible_ai/5
Domain Anchorage in GPT-4: A Computational Linguistic Analysis of Lexicographic Profiling and Its Implications for Unintended Information Dissemination
Hilton Waikoloa Village, Hawaii
Our 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.
https://aisel.aisnet.org/hicss-58/st/safety_trustworthy_and_responsible_ai/5