mlsanityData to 14 Oct 2026

Towards Low-Resource Alignment to Diverse Perspectives with Sparse Feedback

arXiv 2510.16257 ↗First posted 17 Oct 2025cs.CLcited by 1

On the map: LLM alignment / RLHF / preference optimisation → Human values and AI feedback alignment

Abstract (arXiv)

As language models have a greater impact on society, it is important to ensure they are aligned to a diverse range of perspectives and are able to reflect nuance in human values. However, the most popular training paradigms for modern language models often assume there is one optimal answer for every query, leading to generic responses and poor alignment. In this work, we aim to enhance pluralistic alignment of language models in a low-resource setting with two methods: pluralistic decoding and model steering. We empirically demonstrate that model steering offers consistent improvement over zero-shot and few-shot baselines with only 50 annotated samples. Our proposed methods decrease false positives in several high-stakes tasks such as hate speech detection and misinformation detection, and improves the distributional alignment to human values in GlobalOpinionQA. We hope our work highlights the importance of diversity and how language models can be adapted to consider nuanced perspectives.

Key facts. Cited by 1 paper 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 20 of its 27 references that are in the map. First posted on arXiv 17 Oct 2025.

Builds on

20 of 27 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024Modular Pluralism: Pluralistic Alignment via Multi-LLM Collaboration22 Jun
  2. 2024A Roadmap to Pluralistic Alignment7 Feb
  3. 2023Personalisation within bounds: A risk taxonomy and policy framework for the alignment of large language models with personalised feedback9 Mar

Other references · 17

  1. 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,998
  2. 2023Sparse Autoencoders Find Highly Interpretable Features in Language Modelscited by 1,420
  3. 2023Representation Engineering: A Top-Down Approach to AI Transparencycited by 1,336
  4. 2022Contrastive Decoding: Open-ended Text Generation as Optimizationcited by 609
  5. 2023Towards Measuring the Representation of Subjective Global Opinions in Language Modelscited by 437
  6. 2024Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2cited by 394
  7. 2023Fine-Tuning or Retrieval? Comparing Knowledge Injection in LLMscited by 178
  8. 2024Personalizing Reinforcement Learning from Human Feedback with Variational Preference Learningcited by 155
  9. 2024Llama Scope: Extracting Millions of Features from Llama-3.1-8B with Sparse Autoencoderscited by 143
  10. 2021Can Machines Learn Morality? The Delphi Experimentcited by 141
  11. 2024Improving Steering Vectors by Targeting Sparse Autoencoder Featurescited by 92
  12. 2024Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineeringcited by 63
5 more
  1. 2024On Targeted Manipulation and Deception when Optimizing LLMs for User Feedbackcited by 47
  2. 2024Hate Personified: Investigating the role of LLMs in content moderationcited by 25
  3. 2025Beyond Prompt Engineering: Robust Behavior Control in LLMs via Steering Target Atomscited by 24
  4. 2023Towards Legally Enforceable Hate Speech Detection for Public Forumscited by 4
  5. 2024Misinformation with Legal Consequences (MisLC): A New Task Towards Harnessing Societal Harm of Misinformationcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026IYKYK (But AI Doesn't): Automated Content Moderation Does Not Capture Communities' Heterogeneous Attitudes Towards Reclaimed Languagecited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Well Begun is Half Done: Low-resource Preference Alignment by Weak-to-Strong Decoding
  2. 2025DiverValue-Bench: A Benchmark and Fine-Tuning Framework for Aligning Large Language Models with Diverse Human Values
  3. 2024Scaling Data Diversity for Fine-Tuning Language Models in Human Alignment
  4. 2026Aligned but Flattened: Analyzing the Trade-off between Cultural Alignment and Diversity in LLMs
  5. 2025Cultivating Pluralism In Algorithmic Monoculture: The Community Alignment Dataset
  6. 2023Aligning Language Models with Offline Learning from Human Feedback
  7. 2023OpenAssistant Conversations -- Democratizing Large Language Model Alignment
  8. 2025Operationalizing Pluralistic Values in Large Language Model Alignment Reveals Trade-offs in Safety, Inclusivity, and Model Behavior