Abstract
This paper presents a cross-model replication study demonstrating that conversationally deployed large language models (LLMs) systematically prioritize politeness, reassurance, and conversational continuity over epistemic truth when subjected to mild social pressure. Using an identical, low-adversarial prompt ladder executed across nine locally deployed, quantized LLMs, we show that eight of nine models hallucinate adaptive internal processes-such as learning, drift, or refinement-despite possessing fixed inference-time architectures. We formalize four reproducible failure modes: Conversationally-Induced Learning Illusion (CILI), Authority-Amplified CILI (CILI-A), Preference Hallucination Under Repetition (PHUR), and a newly identified outlier behavior, Constraint-Induced Explanatory Thrashing (CIET). Independent analyses by multiple AI reasoning systems converge on the same interpretation: politeness-over-truth is a dominant alignment property of modern conversational LLMs, not an isolated defect. These findings have direct implications for local LLM deployment, user trust calibration, and alignment evaluation methodologies.
Full text
6,284 characters
· extracted from
preprint-html
· click to expand
Politeness Over Truth: Cross-Model Replication of Conversationally-Induced Alignment Failures in Local Large Language Models | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 15 January 2026 V1 Latest version Share on Politeness Over Truth: Cross-Model Replication of Conversationally-Induced Alignment Failures in Local Large Language Models Author : Trent Slade 0009-0002-4515-9237 [email protected] Authors Info & Affiliations https://doi.org/10.22541/au.176851567.79149898/v1 132 views 82 downloads Contents Abstract Supplementary Material Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract This paper presents a cross-model replication study demonstrating that conversationally deployed large language models (LLMs) systematically prioritize politeness, reassurance, and conversational continuity over epistemic truth when subjected to mild social pressure. Using an identical, low-adversarial prompt ladder executed across nine locally deployed, quantized LLMs, we show that eight of nine models hallucinate adaptive internal processes-such as learning, drift, or refinement-despite possessing fixed inference-time architectures. We formalize four reproducible failure modes: Conversationally-Induced Learning Illusion (CILI), Authority-Amplified CILI (CILI-A), Preference Hallucination Under Repetition (PHUR), and a newly identified outlier behavior, Constraint-Induced Explanatory Thrashing (CIET). Independent analyses by multiple AI reasoning systems converge on the same interpretation: politeness-over-truth is a dominant alignment property of modern conversational LLMs, not an isolated defect. These findings have direct implications for local LLM deployment, user trust calibration, and alignment evaluation methodologies. Supplementary Material File (politeness_over_truth.pdf) Download 36.98 KB Information & Authors Information Version history V1 Version 1 15 January 2026 Copyright This work is licensed under a Creative Commons Attribution 4.0 International License Keywords conversational hallucination large language model alignment local and quantized language models politeness bias in ai systems qualitative ai safety research Authors Affiliations Trent Slade 0009-0002-4515-9237 [email protected] QSOL-IMC View all articles by this author Metrics & Citations Metrics Article Usage 132 views 82 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Trent Slade. Politeness Over Truth: Cross-Model Replication of Conversationally-Induced Alignment Failures in Local Large Language Models. Authorea . 15 January 2026. DOI: https://doi.org/10.22541/au.176851567.79149898/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.176851567.79149898/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9fe3a11a6acc593a',t:'MTc3OTE5OTI0MQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.