mlsanityData to 14 Oct 2026

CALMA: A Process for Deriving Context-aligned Axes for Language Model Alignment

arXiv 2507.09060 ↗First posted 11 Jul 2025cs.CYcited by 0

On the map: LLM alignment / RLHF / preference optimisation → LLM preference optimization methods

Abstract (arXiv)

Datasets play a central role in AI governance by enabling both evaluation (measuring capabilities) and alignment (enforcing values) along axes such as helpfulness, harmlessness, toxicity, quality, and more. However, most alignment and evaluation datasets depend on researcher-defined or developer-defined axes curated from non-representative samples. As a result, developers typically benchmark models against broad (often Western-centric) values that overlook the varied contexts of their real-world deployment. Consequently, models trained on such proxies can fail to meet the needs and expectations of diverse user communities within these deployment contexts. To bridge this gap, we introduce CALMA (Context-aligned Axes for Language Model Alignment), a grounded, participatory methodology for eliciting context-relevant axes for evaluation and alignment. In a pilot with two distinct communities, CALMA surfaced novel priorities that are absent from standard benchmarks. Our findings demonstrate the value of evaluation practices based on open-ended and use-case-driven processes. Our work advances the development of pluralistic, transparent, and context-sensitive alignment pipelines.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in LLM preference optimization methods, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 19 of its 32 references that are in the map. First posted on arXiv 11 Jul 2025.

Builds on

19 of 32 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Binary Classifier Optimization for Large Language Model Alignment6 Apr
  2. 2024Arithmetic Control of LLMs for Diverse User Preferences: Directional Preference Alignment with Multi-Objective Rewards28 Feb
  3. 2023Large Language Model Alignment: A Survey26 Sep
  4. 2023Aligning Large Language Models through Synthetic Feedback23 May

Other references · 15

  1. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 3,050
  2. 2022Improving alignment of dialogue agents via targeted human judgementscited by 531
  3. 2021FUDGE: Controlled Text Generation With Future Discriminatorscited by 376
  4. 2022Fine-tuning language models to find agreement among humans with diverse preferencescited by 189
  5. 2024A Roadmap to Pluralistic Alignmentcited by 184
  6. 2021Process for Adapting Language Models to Society (PALMS) with Values-Targeted Datasetscited by 181
  7. 2021On Releasing Annotator-Level Labels and Information in Datasetscited by 126
  8. 2024Collective Constitutional AI: Aligning a Language Model with Public Inputcited by 123
  9. 2023SteerLM: Attribute Conditioned SFT as an (User-Steerable) Alternative to RLHFcited by 111
  10. 2023Generative Social Choicecited by 43
  11. 2023Second Thoughts are Best: Learning to Re-Align With Human Values from Text Editscited by 27
  12. 2023Indian-BhED: A Dataset for Measuring India-Centric Biases in Large Language Modelscited by 18
3 more
  1. 2024Wikibench: Community-Driven Data Curation for AI Evaluation on Wikipediacited by 18
  2. 2025Pitfalls of Evidence-Based AI Policycited by 11
  3. 2024LocalValueBench: A Collaboratively Built and Extensible Benchmark for Evaluating Localized Value Alignment and Ethical Safety in Large Language Modelscited by 7

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Alignment and Safety in Large Language Models: Safety Mechanisms, Training Paradigms, and Emerging Challenges
  2. 2024Aligner: Efficient Alignment by Learning to Correct
  3. 2026SpecAlign: Efficient Specification-Grounded Alignment of Large Language Models via Synthetic Data
  4. 2025Dynamic Alignment for Collective Agency: Toward a Scalable Self-Improving Framework for Open-Ended LLM Alignment
  5. 2025Decoding Alignment: A Critical Survey of LLM Development Initiatives through Value-setting and Data-centric Lens
  6. 2026Pressure Reveals Character: Behavioural Alignment Evaluation at Depth
  7. 2026AlignTune: Modular Toolkit for Post-Training Alignment of Large Language Models
  8. 2025P-Aligner: Enabling Pre-Alignment of Language Models via Principled Instruction Synthesis