LLMs Can Better Capture Human Judgments--With the Right Prompts

• URL: http://arxiv.org/abs/2606.12754v1
• Date: Wed, 10 Jun 2026 23:42:22 GMT
• Title: LLMs Can Better Capture Human Judgments--With the Right Prompts
• Abstract summary: Two commonly stated limitations are that large language models fail to capture full distributions of responses, and their judgments are unstable across wording variations.<n>We show how simple elicitation techniques help improve AI-human alignment.
• Score: 7.957808148711568
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Are large language models (LLMs) bad at capturing human judgment? Two commonly stated limitations are that LLMs fail to capture full distributions of responses, and that their judgments are unstable across wording variations. We demonstrate simple prompting strategies that mitigate these limitations. Across two datasets--a U.S.-representative set of 144 moral scenarios and 38 moral beliefs from the International Social Survey Programme's Family and Changing Gender Roles module covering 32 countries--we show how simple elicitation techniques help improve AI-human alignment. First, prompting models to report standard deviations and response proportions recovers the full range of human responses better than common strategies. Second, ensuring scenarios are clear to human participants--as reflected in human confusion ratings--boosts model alignment, and LLMs can track human confusion ratings. At the same time, we find that LLMs' estimates of their own error are poorly calibrated, though they can predict human variability relatively well. These results suggest that asking better questions to LLMs can yield better answers.

Related papers

• Do LLMs Change Their Minds Like Humans? Diagnosing Human--LLM Divergence in Single-Turn Persuasion Judgments [10.717755637740948]
  Large language models (LLMs) are increasingly deployed as proxies for human participants in social simulations.<n>We conduct a systematic comparison using a naturally occurring online persuasion corpus.<n>Our results show that LLMs achieve only slight agreement with humans.
  arXiv  Detail & Related papers  (2026-08-30T14:08:06Z)
• Augmenting Human Evaluation with LLM Judges: How Many Human Reviews Do You Need? [1.111977509278551]
  Large language models (LLMs) are increasingly used as automated evaluators of AI systems, including in high-stakes applications.<n> Expert human ratings are costly and difficult to scale, whereas LLM ratings can be produced quickly at low cost.<n>This paper shifts the role of the LLM judge from substitutive to auxiliary, and formulates the LLM-as-a-judge paradigm as one of augmenting human evaluation.
  arXiv  Detail & Related papers  (2026-05-08T17:13:08Z)
• Blind to the Human Touch: Overlap Bias in LLM-Based Summary Evaluation [89.52571224447111]
  Large language model (LLM) judges have often been used alongside traditional, algorithm-based metrics for tasks like summarization.<n>We provide an LLM judge bias analysis as a function of overlap with human-written responses in the domain of summarization.
  arXiv  Detail & Related papers  (2026-02-07T19:39:28Z)
• Do LLMs Align Human Values Regarding Social Biases? Judging and Explaining Social Biases with LLMs [24.53996114318076]
  Large language models (LLMs) can lead to undesired consequences when misaligned with human values.<n>Previous studies have revealed the misalignment of LLMs with human values using expert-designed or agent-based emulated bias scenarios.<n>In this study, we investigate the alignment of LLMs with human values regarding social biases (HVSB) in different types of bias scenarios.
  arXiv  Detail & Related papers  (2025-09-17T09:58:28Z)
• Large Language Model Hacking: Quantifying the Hidden Risks of Using LLMs for Text Annotation [66.84286617519258]
  Large language models are transforming social science research by enabling the automation of labor-intensive tasks like data annotation and text analysis.<n>Such variation can introduce systematic biases and random errors, which propagate to downstream analyses and cause Type I (false positive), Type II (false negative), Type S (wrong sign), or Type M (exaggerated effect) errors.<n>We find that intentional LLM hacking is strikingly simple. By replicating 37 data annotation tasks from 21 published social science studies, we show that, with just a handful of prompt paraphrases, virtually anything can be presented as statistically significant.
  arXiv  Detail & Related papers  (2025-09-10T17:58:53Z)
• Arbiters of Ambivalence: Challenges of Using LLMs in No-Consensus Tasks [52.098988739649705]
  This study examines the biases and limitations of LLMs in three roles: answer generator, judge, and debater.<n>We develop a no-consensus'' benchmark by curating examples that encompass a variety of a priori ambivalent scenarios.<n>Our results show that while LLMs can provide nuanced assessments when generating open-ended answers, they tend to take a stance on no-consensus topics when employed as judges or debaters.
  arXiv  Detail & Related papers  (2025-05-28T01:31:54Z)
• Large Language Models are overconfident and amplify human bias [1.014221700787766]
  We evaluate whether large language models (LLMs) inherit one of the most widespread human biases: overconfidence.<n>All five LLMs we study are overconfident: they overestimate the probability that their answer is correct between 20% and 60%.<n>Humans have accuracy similar to the more advanced LLMs, but far lower overconfidence.
  arXiv  Detail & Related papers  (2025-05-04T15:21:34Z)
• Scaling Data Diversity for Fine-Tuning Language Models in Human Alignment [84.32768080422349]
  Alignment with human preference prevents large language models from generating misleading or toxic content. We propose a new formulation of prompt diversity, implying a linear correlation with the final performance of LLMs after fine-tuning.
  arXiv  Detail & Related papers  (2024-03-17T07:08:55Z)
• Are You Sure? Challenging LLMs Leads to Performance Drops in The FlipFlop Experiment [82.60594940370919]
  We propose the FlipFlop experiment to study the multi-turn behavior of Large Language Models (LLMs) We show that models flip their answers on average 46% of the time and that all models see a deterioration of accuracy between their first and final prediction, with an average drop of 17% (the FlipFlop effect) We conduct finetuning experiments on an open-source LLM and find that finetuning on synthetically created data can mitigate - reducing performance deterioration by 60% - but not resolve sycophantic behavior entirely.
  arXiv  Detail & Related papers  (2023-11-14T23:40:22Z)
• Do LLMs exhibit human-like response biases? A case study in survey design [66.1850490474361]
  We investigate the extent to which large language models (LLMs) reflect human response biases, if at all. We design a dataset and framework to evaluate whether LLMs exhibit human-like response biases in survey questionnaires. Our comprehensive evaluation of nine models shows that popular open and commercial LLMs generally fail to reflect human-like behavior.
  arXiv  Detail & Related papers  (2023-11-07T15:40:43Z)

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.