The Rise of Verbal Tics in Large Language Models: A Systematic Analysis Across Frontier Models

TL;DR

This paper systematically analyzes the prevalence of verbal tics in eight state-of-the-art large language models, introducing the Verbal Tic Index and revealing significant inter-model and cross-lingual variations.

Contribution

It presents a novel evaluation framework and the Verbal Tic Index to quantify verbal tics, highlighting their correlation with model alignment and naturalness across multiple models and languages.

Findings

Gemini 3.1 Pro has the highest Verbal Tic Index (0.590)

DeepSeek V3.2 has the lowest Verbal Tic Index (0.295)

Verbal tics increase over multi-turn conversations and are more common in subjective tasks.

Abstract

As Large Language Models (LLMs) continue to evolve through alignment techniques such as Reinforcement Learning from Human Feedback (RLHF) and Constitutional AI, a growing and increasingly conspicuous phenomenon has emerged: the proliferation of verbal tics, repetitive, formulaic linguistic patterns that pervade model outputs. These range from sycophantic openers (That's a great question!, Awesome!) to pseudo-empathetic affirmations (I completely understand your concern, I'm right here to catch you) and overused vocabulary (delve, tapestry, nuanced). In this paper, we present a systematic analysis of the verbal tic phenomenon across eight state-of-the-art LLMs: GPT-5.4, Claude Opus 4.7, Gemini 3.1 Pro, Grok 4.2, Doubao-Seed-2.0-pro, Kimi K2.5, DeepSeek V3.2, and MiMo-V2-Pro. Utilizing a custom evaluation framework for standardized API-based evaluation, we assess 10,000 prompts across 10 task categories in both English and Chinese, yielding 160,000 model responses. We introduce the Verbal Tic Index (VTI), a composite metric quantifying tic prevalence, and analyze its correlation with sycophancy, lexical diversity, and human-perceived naturalness. Our findings reveal significant inter-model variation: Gemini 3.1 Pro exhibits the highest VTI (0.590), while DeepSeek V3.2 achieves the lowest (0.295). We further demonstrate that verbal tics accumulate over multi-turn conversations, are amplified in subjective tasks, and show distinct cross-lingual patterns. Human evaluation (N = 120) confirms a strong inverse relationship between sycophancy and perceived naturalness (r = -0.87, p < 0.001). These results underscore the alignment tax of current training paradigms and highlight the urgent need for more authentic human-AI interaction frameworks.