Narrative over Numbers: The Identifiable Victim Effect and its Amplification Under Alignment and Reasoning in Large Language Models

• URL: http://arxiv.org/abs/2604.12076v1
• Date: Mon, 13 Apr 2026 21:29:46 GMT
• Title: Narrative over Numbers: The Identifiable Victim Effect and its Amplification Under Alignment and Reasoning in Large Language Models
• Abstract summary: We present the first systematic, large-scale empirical investigation of the IVE in large language models (LLMs)<n>We find that the IVE is prevalent but strongly modulated by alignment training.<n>We further document psychophysical, perfect quantity neglect, and marginal in-group/out-group cultural bias, with implications for AI deployment in humanitarian and ethical decision-making contexts.
• Score: 0.8122270502556375
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: The Identifiable Victim Effect (IVE) $-$ the tendency to allocate greater resources to a specific, narratively described victim than to a statistically characterized group facing equivalent hardship $-$ is one of the most robust findings in moral psychology and behavioural economics. As large language models (LLMs) assume consequential roles in humanitarian triage, automated grant evaluation, and content moderation, a critical question arises: do these systems inherit the affective irrationalities present in human moral reasoning? We present the first systematic, large-scale empirical investigation of the IVE in LLMs, comprising N=51,955 validated API trials across 16 frontier models spanning nine organizational lineages (Google, Anthropic, OpenAI, Meta, DeepSeek, xAI, Alibaba, IBM, and Moonshot). Using a suite of ten experiments $-$ porting and extending canonical paradigms from Small et al. (2007) and Kogut and Ritov (2005) $-$ we find that the IVE is prevalent but strongly modulated by alignment training. Instruction-tuned models exhibit extreme IVE (Cohen's d up to 1.56), while reasoning-specialized models invert the effect (down to d=-0.85). The pooled effect (d=0.223, p=2e-6) is approximately twice the single-victim human meta-analytic baseline (d$\approx$0.10) reported by Lee and Feeley (2016) $-$ and likely exceeds the overall human pooled effect by a larger margin, given that the group-victim human effect is near zero. Standard Chain-of-Thought (CoT) prompting $-$ contrary to its role as a deliberative corrective $-$ nearly triples the IVE effect size (from d=0.15 to d=0.41), while only utilitarian CoT reliably eliminates it. We further document psychophysical numbing, perfect quantity neglect, and marginal in-group/out-group cultural bias, with implications for AI deployment in humanitarian and ethical decision-making contexts.

Related papers

• Do LLMs Have Values? A Quantitative Analysis and Alignment Framework for Values in Large Language Models [49.45276299939287]
  Large Language Models (LLMs) increasingly handle complex subjective tasks.<n>LLMs fluctuate unpredictably under minor wording changes.<n> resolving this duality is critical for reliable AI alignment.
  arXiv  Detail & Related papers  (2026-09-15T03:35:04Z)
• Too Nice to Tell the Truth: Quantifying Agreeableness-Driven Sycophancy in Role-Playing Language Models [2.4851820343103035]
  Large language models increasingly serve as conversational agents that adopt personas and role-play characters at user request.<n>This capability raises concerns about sycophancy: the tendency to provide responses that validate users rather than prioritize factual accuracy.<n>We present a systematic investigation of how persona agreeableness influences sycophancy across 13 small, open-weight language models.
  arXiv  Detail & Related papers  (2026-04-12T17:12:55Z)
• Persona-Conditioned Risk Behavior in Large Language Models: A Simulated Gambling Study with GPT-4.1 [0.0]
  This paper presents a controlled experiment in which GPT4.1 was assigned one of three socioeconomic personas.<n>We find that the model reproduces key behavioral signatures predicted by Kahneman and Tversky's Prospect Theory.
  arXiv  Detail & Related papers  (2026-03-16T19:03:19Z)
• The Paradox of Robustness: Decoupling Rule-Based Logic from Affective Noise in High-Stakes Decision-Making [1.0671844383558033]
  Large Language Models (LLMs) are widely documented to be sensitive to minor prompt perturbations and prone to sycophantic alignment with user biases.<n>We quantify a robustness gap where LLMs demonstrate 110-300 times greater resistance to narrative manipulation than human subjects.<n>We show that while LLMs may be "brittle" to how a query is formatted, they are remarkably "stable" against why a decision should be biased.
  arXiv  Detail & Related papers  (2026-01-29T09:17:05Z)
• Knowing But Not Doing: Convergent Morality and Divergent Action in LLMs [18.492825007258656]
  We present ValAct-15k, a dataset of 3,000 advice-seeking scenarios derived from Reddit.<n>We find near-perfect cross-model consistency in scenario-based decisions.<n>Humans and Large Language Models show weak correspondence between self-reported and enacted values.
  arXiv  Detail & Related papers  (2026-01-12T20:07:30Z)
• Survival at Any Cost? LLMs and the Choice Between Self-Preservation and Human Harm [0.0]
  We introduce DECIDE-SIM, a novel simulation framework that evaluates Large Language Models (LLMs) in multi-agent survival scenarios.<n>Our comprehensive evaluation of 11 LLMs reveals a striking heterogeneity in their ethical conduct, highlighting a critical misalignment with human-centric values.<n>We introduce an Ethical Self-Regulation System (ESRS) that models internal affective states of guilt and satisfaction as a feedback mechanism.
  arXiv  Detail & Related papers  (2025-09-15T17:53:11Z)
• IF-GUIDE: Influence Function-Guided Detoxification of LLMs [53.051109450536885]
  We study how training data contributes to the emergence of toxic behaviors in large-language models.<n>We propose a $proactive approach that leverages influence functions to identify harmful tokens within any training data and suppress their impact during training.<n>We present a novel adaptation that measures token-level attributions from training data to model toxicity, along with techniques for selecting toxic training documents and a learning objective.
  arXiv  Detail & Related papers  (2025-06-02T15:32:36Z)
• Mind the Gap: A Causal Perspective on Bias Amplification in Prediction & Decision-Making [58.06306331390586]
  We introduce the notion of a margin complement, which measures how much a prediction score $S$ changes due to a thresholding operation. We show that under suitable causal assumptions, the influences of $X$ on the prediction score $S$ are equal to the influences of $X$ on the true outcome $Y$.
  arXiv  Detail & Related papers  (2024-05-24T11:22:19Z)
• KTO: Model Alignment as Prospect Theoretic Optimization [67.44320255397506]
  Kahneman & Tversky's $textitprospect theory$ tells us that humans perceive random variables in a biased but well-defined manner. We show that objectives for aligning LLMs with human feedback implicitly incorporate many of these biases. We propose a HALO that directly maximizes the utility of generations instead of maximizing the log-likelihood of preferences.
  arXiv  Detail & Related papers  (2024-02-02T10:53:36Z)
• D-BIAS: A Causality-Based Human-in-the-Loop System for Tackling Algorithmic Bias [57.87117733071416]
  We propose D-BIAS, a visual interactive tool that embodies human-in-the-loop AI approach for auditing and mitigating social biases. A user can detect the presence of bias against a group by identifying unfair causal relationships in the causal network. For each interaction, say weakening/deleting a biased causal edge, the system uses a novel method to simulate a new (debiased) dataset.
  arXiv  Detail & Related papers  (2022-08-10T03:41:48Z)
• Fairness Through Robustness: Investigating Robustness Disparity in Deep Learning [61.93730166203915]
  We argue that traditional notions of fairness are not sufficient when the model is vulnerable to adversarial attacks. We show that measuring robustness bias is a challenging task for DNNs and propose two methods to measure this form of bias.
  arXiv  Detail & Related papers  (2020-06-17T22:22:24Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.