Metraipolis › archive › dc_eeb0142d170b
Successive Training Stages and Large Language Model Persuasion: Effects of Misalignment, Supervised Fine-Tuning, and Preference Optimization — arXiv:2610.09964
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.09964
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T11:09:40.629Z
- SHA-256 of the archived text
9c4ccba29879e5c87f895e1d43da766a5f701be8e55d745d99958efd7d58836a- Stable URL of this page
https://metraipolis.com/documento/dc_eeb0142d170b
How to cite
Successive Training Stages and Large Language Model Persuasion: Effects of Misalignment, Supervised Fine-Tuning, and Preference Optimization — arXiv:2610.09964 — arXiv (en), https://arxiv.org/abs/2610.09964 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_eeb0142d170b (Metraipolis archive dc_eeb0142d170b, sha256 9c4ccba29879e5c87f895e1d43da766a5f701be8e55d745d99958efd7d58836a)
BibTeX
@misc{dceeb0142d170b,
title = {Successive Training Stages and Large Language Model Persuasion: Effects of Misalignment, Supervised Fine-Tuning, and Preference Optimization — arXiv:2610.09964},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.09964},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_eeb0142d170b; sha256 9c4ccba29879e5c87f895e1d43da766a5f701be8e55d745d99958efd7d58836a},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.09964. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Large language models (LLMs) can be tuned to influence human attitudes, yet the respective contributions of successive post-training stages remain un-clear. This study examines how three successive training stages affect LLM persuasiveness: (1) misalignment through supervised fine-tuning (SFT) on conspiracy data, (2) additional persuasive SFT on argumentative data, and (3) Identity Preference Optimization (IPO), a preference-optimization method. A total of 835 participants recruited on Prolific were randomly assigned to five between-subject conditions (neutral text, conspiracy-trained model, persuasion-trained model, preference-optimized model, and GPT-4) and were exposed to texts on 10 divisive political issues, personalized from their individual profiles in all model conditions. Attitude change was measured as the difference between pre- and post-exposure positions on continuous Likert scales and analyzed with an analysis of covariance (ANCOVA). A significant condition x baseline-attitude interaction, F (4, 825) = 5.33, p < .001, indicated that training effects depended on participants' initial attitudes. Persuasive SFT produced greater attitude change than conspiracy training alone, d = 0.30, whereas IPO provided no additional benefit, d = 0.03, and GPT-4 did not differ from neutral text, d = --0.01. These results show that targeted supervised training on persuasive data increases LLM persuasiveness, whereas preference optimization yields no significant gains beyond it. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI) ; Computers and Society (cs.CY). Pagina del lavoro: https://arxiv.org/abs/2610.09964 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-08T10:14:05.968Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).