# Dissecting Human and LLM Preferences

arXiv: https://arxiv.org/abs/2402.11296 · HTML: https://mlsanity.com/paper/2402.11296/

## Key facts

- Cited by 17 papers in the map (data through 14 Oct 2026).
- Placed in Human values and AI feedback alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 25 of its 40 references that are in the map.
- First posted on arXiv 17 Feb 2024.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Human values and AI feedback alignment](https://mlsanity.com/alignment/human-values-and-ai-feedback-alignment/)

## Builds on (most cited first, 25 of 25)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [Qwen Technical Report](https://mlsanity.com/paper/2309.16609/) (2023)
- [Fine-Tuning Language Models from Human Preferences](https://mlsanity.com/paper/1909.08593/) (2019)
- [WizardLM: Empowering large pre-trained language models to follow complex instructions](https://mlsanity.com/paper/2304.12244/) (2023)
- [Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting](https://mlsanity.com/paper/2305.04388/) (2023)
- [LIMA: Less Is More for Alignment](https://mlsanity.com/paper/2305.11206/) (2023)
- [Scaling Laws for Reward Model Overoptimization](https://mlsanity.com/paper/2210.10760/) (2022)
- [Large Language Models are not Fair Evaluators](https://mlsanity.com/paper/2305.17926/) (2023)
- [Towards Understanding Sycophancy in Language Models](https://mlsanity.com/paper/2310.13548/) (2023)
- [Discovering Language Model Behaviors with Model-Written Evaluations](https://mlsanity.com/paper/2212.09251/) (2022)
- [Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2307.15217/) (2023)
- [AlpacaFarm: A Simulation Framework for Methods that Learn from Human Feedback](https://mlsanity.com/paper/2305.14387/) (2023)
- [RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedback](https://mlsanity.com/paper/2309.00267/) (2023)
- [Zephyr: Direct Distillation of LM Alignment](https://mlsanity.com/paper/2310.16944/) (2023)
- [Self-critiquing models for assisting human evaluators](https://mlsanity.com/paper/2206.05802/) (2022)
- [A Long Way to Go: Investigating Length Correlations in RLHF](https://mlsanity.com/paper/2310.03716/) (2023)
- [The False Promise of Imitating Proprietary LLMs](https://mlsanity.com/paper/2305.15717/) (2023)
- [Camels in a Changing Climate: Enhancing LM Adaptation with Tulu 2](https://mlsanity.com/paper/2311.10702/) (2023)
- [Composable Effects for Flexible and Accelerated Probabilistic Programming in NumPyro](https://mlsanity.com/paper/1912.11554/) (2019)
- [Human Feedback is not Gold Standard](https://mlsanity.com/paper/2309.16349/) (2023)
- [Modeling and mitigating human annotation errors to design efficient stream processing systems with human-in-the-loop machine learning](https://mlsanity.com/paper/2007.03177/) (2020)

## Built on by (most cited first, 17 of 17)

- [Do LLMs Recognize Your Preferences? Evaluating Personalized Preference Following in LLMs](https://mlsanity.com/paper/2502.09597/) (2025)
- [Aligning to Thousands of Preferences via System Message Generalization](https://mlsanity.com/paper/2405.17977/) (2024)
- [Neither Valid nor Reliable? Investigating the Use of LLMs as Judges](https://mlsanity.com/paper/2508.18076/) (2025)
- [Towards Automating Data Access Permissions in AI Agents](https://mlsanity.com/paper/2511.17959/) (2025)
- [AI Alignment at Your Discretion](https://mlsanity.com/paper/2502.10441/) (2025)
- [I2-NeRF: Learning Neural Radiance Fields Under Physically-Grounded Media Interactions](https://mlsanity.com/paper/2510.22161/) (2025)
- [KoBALT: Korean Benchmark For Advanced Linguistic Tasks](https://mlsanity.com/paper/2505.16125/) (2025)
- [Socratic Reasoning Improves Positive Text Rewriting](https://mlsanity.com/paper/2403.03029/) (2024)
- [More Code, Less Reuse: Investigating Code Quality and Reviewer Sentiment towards AI-generated Pull Requests](https://mlsanity.com/paper/2601.21276/) (2026)
- [Can LLMs Make (Personalized) Access Control Decisions?](https://mlsanity.com/paper/2511.20284/) (2025)
- [Dataset Featurization: Uncovering Natural Language Features through Unsupervised Data Reconstruction](https://mlsanity.com/paper/2502.17541/) (2025)
- [Unpacking Human Preference for LLMs: Demographically Aware Evaluation with the HUMAINE Framework](https://mlsanity.com/paper/2603.04409/) (2026)
- [Janus: a Playground for User-Involved Agentic Permission Management](https://mlsanity.com/paper/2607.01510/) (2026)
- [Aligning Deep Implicit Preferences by Learning to Reason Defensively](https://mlsanity.com/paper/2510.11194/) (2025)
- [SynCast: Synergizing Contradictions in Precipitation Nowcasting via Diffusion Sequential Preference Optimization](https://mlsanity.com/paper/2510.21847/) (2025)
- [Human Preferences for Constructive Interactions in Language Model Alignment](https://mlsanity.com/paper/2503.16480/) (2025)
- [Spectral Souping: A Unified Framework for Online Preference Alignment](https://mlsanity.com/paper/2605.20408/) (2026)

## Abstract (arXiv)

As a relative quality comparison of model responses, human and Large Language Model (LLM) preferences serve as common alignment goals in model fine-tuning and criteria in evaluation. Yet, these preferences merely reflect broad tendencies, resulting in less explainable and controllable models with potential safety risks. In this work, we dissect the preferences of human and 32 different LLMs to understand their quantitative composition, using annotations from real-world user-model conversations for a fine-grained, scenario-wise analysis. We find that humans are less sensitive to errors, favor responses that support their stances, and show clear dislike when models admit their limits. On the contrary, advanced LLMs like GPT-4-Turbo emphasize correctness, clarity, and harmlessness more. Additionally, LLMs of similar sizes tend to exhibit similar preferences, regardless of their training methods, and fine-tuning for alignment does not significantly alter the preferences of pretrained-only LLMs. Finally, we show that preference-based evaluation can be intentionally manipulated. In both training-free and training-based settings, aligning a model with the preferences of judges boosts scores, while injecting the least preferred properties lowers them. This results in notable score shifts: up to 0.59 on MT-Bench (1-10 scale) and 31.94 on AlpacaEval 2.0 (0-100 scale), highlighting the significant impact of this strategic adaptation. Interactive Demo: https://huggingface.co/spaces/GAIR/Preference-Dissection-Visualization Dataset: https://huggingface.co/datasets/GAIR/preference-dissection Code: https://github.com/GAIR-NLP/Preference-Dissection