# Peering Through Preferences: Unraveling Feedback Acquisition for Aligning Large Language Models

arXiv: https://arxiv.org/abs/2308.15812 · HTML: https://mlsanity.com/paper/2308.15812/

## Key facts

- Cited by 23 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 44 of its 63 references that are in the map.
- First posted on arXiv 30 Aug 2023.

## 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, 30 of 44)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [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)
- [Self-Consistency Improves Chain of Thought Reasoning in Language Models](https://mlsanity.com/paper/2203.11171/) (2022)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Let's Verify Step by Step](https://mlsanity.com/paper/2305.20050/) (2023)
- [Finetuned Language Models Are Zero-Shot Learners](https://mlsanity.com/paper/2109.01652/) (2021)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [OPT: Open Pre-trained Transformer Language Models](https://mlsanity.com/paper/2205.01068/) (2022)
- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [Constitutional AI: Harmlessness from AI Feedback](https://mlsanity.com/paper/2212.08073/) (2022)
- [Training Compute-Optimal Large Language Models](https://mlsanity.com/paper/2203.15556/) (2022)
- [Self-Instruct: Aligning Language Models with Self-Generated Instructions](https://mlsanity.com/paper/2212.10560/) (2022)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [Language Models (Mostly) Know What They Know](https://mlsanity.com/paper/2207.05221/) (2022)
- [WebGPT: Browser-assisted question-answering with human feedback](https://mlsanity.com/paper/2112.09332/) (2021)
- [WizardLM: Empowering large pre-trained language models to follow complex instructions](https://mlsanity.com/paper/2304.12244/) (2023)
- [A General Language Assistant as a Laboratory for Alignment](https://mlsanity.com/paper/2112.00861/) (2021)
- [ChatGPT Outperforms Crowd-Workers for Text-Annotation Tasks](https://mlsanity.com/paper/2303.15056/) (2023)
- [Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks](https://mlsanity.com/paper/2204.07705/) (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)
- [Instruction Tuning with GPT-4](https://mlsanity.com/paper/2304.03277/) (2023)
- [RAFT: Reward rAnked FineTuning for Generative Foundation Model Alignment](https://mlsanity.com/paper/2304.06767/) (2023)

## Built on by (most cited first, 23 of 23)

- [Evaluating Large Language Models at Evaluating Instruction Following](https://mlsanity.com/paper/2310.07641/) (2023)
- [RLHF Workflow: From Reward Modeling to Online RLHF](https://mlsanity.com/paper/2405.07863/) (2024)
- [Arithmetic Control of LLMs for Diverse User Preferences: Directional Preference Alignment with Multi-Objective Rewards](https://mlsanity.com/paper/2402.18571/) (2024)
- [Survey of Bias In Text-to-Image Generation: Definition, Evaluation, and Mitigation](https://mlsanity.com/paper/2404.01030/) (2024)
- [Group Preference Optimization: Few-Shot Alignment of Large Language Models](https://mlsanity.com/paper/2310.11523/) (2023)
- [A Comprehensive Survey of Reward Models: Taxonomy, Applications, Challenges, and Future](https://mlsanity.com/paper/2504.12328/) (2025)
- [RLHF Can Speak Many Languages: Unlocking Multilingual Preference Optimization for LLMs](https://mlsanity.com/paper/2407.02552/) (2024)
- [The History and Risks of Reinforcement Learning and Human Feedback](https://mlsanity.com/paper/2310.13595/) (2023)
- [Inverse Constitutional AI: Compressing Preferences into Principles](https://mlsanity.com/paper/2406.06560/) (2024)
- [Pairwise or Pointwise? Evaluating Feedback Protocols for Bias in LLM-Based Evaluation](https://mlsanity.com/paper/2504.14716/) (2025)
- [Exploring the impact of low-rank adaptation on the performance, efficiency, and regularization of RLHF](https://mlsanity.com/paper/2309.09055/) (2023)
- [Direct Alignment with Heterogeneous Preferences](https://mlsanity.com/paper/2502.16320/) (2025)
- [Position: Towards Bidirectional Human-AI Alignment](https://mlsanity.com/paper/2406.09264/) (2024)
- [GTPO and GRPO-S: Token and Sequence-Level Reward Shaping with Policy Entropy](https://mlsanity.com/paper/2508.04349/) (2025)
- [Reasons to Reject? Aligning Language Models with Judgments](https://mlsanity.com/paper/2312.14591/) (2023)
- [Comparing Bad Apples to Good Oranges: Aligning Large Language Models via Joint Preference Optimization](https://mlsanity.com/paper/2404.00530/) (2024)
- [Flattery, Fluff, and Fog: Diagnosing and Mitigating Idiosyncratic Biases in Preference Models](https://mlsanity.com/paper/2506.05339/) (2025)
- [Varying Shades of Wrong: Aligning LLMs with Wrong Answers Only](https://mlsanity.com/paper/2410.11055/) (2024)
- [CURE-Med: Curriculum-Informed Reinforcement Learning for Multilingual Medical Reasoning](https://mlsanity.com/paper/2601.13262/) (2026)
- [Multilingual Arbitrage: Optimizing Data Pools to Accelerate Multilingual Progress](https://mlsanity.com/paper/2408.14960/) (2024)
- [LASeR: Learning to Adaptively Select Reward Models with Multi-Armed Bandits](https://mlsanity.com/paper/2410.01735/) (2024)
- [Epitome: Pioneering an Experimental Platform for AI-Social Science Integration](https://mlsanity.com/paper/2507.01061/) (2025)
- [Preferences Order, Ratings Anchor: From Fused Expert Aesthetic Ground Truth to Self-Distillation](https://mlsanity.com/paper/2605.19776/) (2026)

## Abstract (arXiv)

Aligning large language models (LLMs) with human values and intents critically involves the use of human or AI feedback. While dense feedback annotations are expensive to acquire and integrate, sparse feedback presents a structural design choice between ratings (e.g., score Response A on a scale of 1-7) and rankings (e.g., is Response A better than Response B?). In this work, we analyze the effect of this design choice for the alignment and evaluation of LLMs. We uncover an inconsistency problem wherein the preferences inferred from ratings and rankings significantly disagree 60% for both human and AI annotators. Our subsequent analysis identifies various facets of annotator biases that explain this phenomena, such as human annotators would rate denser responses higher while preferring accuracy during pairwise judgments. To our surprise, we also observe that the choice of feedback protocol also has a significant effect on the evaluation of aligned LLMs. In particular, we find that LLMs that leverage rankings data for alignment (say model X) are preferred over those that leverage ratings data (say model Y), with a rank-based evaluation protocol (is X/Y's response better than reference response?) but not with a rating-based evaluation protocol (score Rank X/Y's response on a scale of 1-7). Our findings thus shed light on critical gaps in methods for evaluating the real-world utility of language models and their strong dependence on the feedback protocol used for alignment. Our code and data are available at https://github.com/Hritikbansal/sparse_feedback.