# EMBRACE: Shaping Inclusive Opinion Representation by Aligning Implicit Conversations with Social Norms

arXiv: https://arxiv.org/abs/2507.20264 · HTML: https://mlsanity.com/paper/2507.20264/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 50 references that are in the map.
- First posted on arXiv 27 Jul 2025.

## Builds on (most cited first, 24 of 24)

- [WildChat: 1M ChatGPT Interaction Logs in the Wild](https://mlsanity.com/paper/2405.01470/) (2024)
- [Social Bias Frames: Reasoning about Social and Power Implications of Language](https://mlsanity.com/paper/1911.03891/) (2019)
- [ToxicChat: Unveiling Hidden Challenges of Toxicity Detection in Real-World User-AI Conversation](https://mlsanity.com/paper/2310.17389/) (2023)
- [Latent Hatred: A Benchmark for Understanding Implicit Hate Speech](https://mlsanity.com/paper/2109.05322/) (2021)
- [Modular Pluralism: Pluralistic Alignment via Multi-LLM Collaboration](https://mlsanity.com/paper/2406.15951/) (2024)
- [Measuring Sycophancy of Language Models in Multi-turn Dialogues](https://mlsanity.com/paper/2505.23840/) (2025)
- [Stance Detection on Social Media: State of the Art and Trends](https://mlsanity.com/paper/2006.03644/) (2020)
- [Unintended Impacts of LLM Alignment on Global Representation](https://mlsanity.com/paper/2402.15018/) (2024)
- [Towards Implicit Bias Detection and Mitigation in Multi-Agent LLM Interactions](https://mlsanity.com/paper/2410.02584/) (2024)
- [Human-Machine Collaboration Approaches to Build a Dialogue Dataset for Hate Speech Countering](https://mlsanity.com/paper/2211.03433/) (2022)
- [From Distributional to Overton Pluralism: Investigating Large Language Model Alignment](https://mlsanity.com/paper/2406.17692/) (2024)
- [Unmasking Implicit Bias: Evaluating Persona-Prompted LLM Responses in Power-Disparate Social Scenarios](https://mlsanity.com/paper/2503.01532/) (2025)
- [The Role of Conversation Context for Sarcasm Detection in Online Interactions](https://mlsanity.com/paper/1707.06226/) (2017)
- [Aligned but Blind: Alignment Increases Implicit Bias by Reducing Awareness of Race](https://mlsanity.com/paper/2506.00253/) (2025)
- [Chain-of-Thought Embeddings for Stance Detection on Social Media](https://mlsanity.com/paper/2310.19750/) (2023)
- [Labeling Messages as AI-Generated Does Not Reduce Their Persuasive Effects](https://mlsanity.com/paper/2504.09865/) (2025)
- [Subtle Biases Need Subtler Measures: Dual Metrics for Evaluating Representative and Affinity Bias in Large Language Models](https://mlsanity.com/paper/2405.14555/) (2024)
- [Aligning Large Language Models with Implicit Preferences from User-Generated Content](https://mlsanity.com/paper/2506.04463/) (2025)
- [Evaluating Implicit Bias in Large Language Models by Attacking From a Psychometric Perspective](https://mlsanity.com/paper/2406.14023/) (2024)
- [Superlatives in Context: Modeling the Implicit Semantics of Superlatives](https://mlsanity.com/paper/2405.20967/) (2024)
- [What's the Difference? Supporting Users in Identifying the Effects of Prompt and Model Changes Through Token Patterns](https://mlsanity.com/paper/2504.15815/) (2025)
- [Covert Bias: The Severity of Social Views' Unalignment in Language Models Towards Implicit and Explicit Opinion](https://mlsanity.com/paper/2408.08212/) (2024)
- [Guiding Computational Stance Detection with Expanded Stance Triangle Framework](https://mlsanity.com/paper/2305.19845/) (2023)
- [USDC: A Dataset of underlineUser underlineStance and underlineDogmatism in Long underlineConversations](https://mlsanity.com/paper/2406.16833/) (2024)

## Built on by (most cited first, 1 of 1)

- [Same Lesson, Different Story: Cross-Lingual Reconstruction of Cultural Narratives in Large Language Models](https://mlsanity.com/paper/2606.24610/) (2026)

## Abstract (arXiv)

Shaping inclusive representations that embrace diversity and ensure fair participation and reflections of values is at the core of many conversation-based models. However, many existing methods rely on surface inclusion using mention of user demographics or behavioral attributes of social groups. Such methods overlook the nuanced, implicit expression of opinion embedded in conversations. Furthermore, the over-reliance on overt cues can exacerbate misalignment and reinforce harmful or stereotypical representations in model outputs. Thus, we took a step back and recognized that equitable inclusion needs to account for the implicit expression of opinion and use the stance of responses to validate the normative alignment. This study aims to evaluate how opinions are represented in NLP or computational models by introducing an alignment evaluation framework that foregrounds implicit, often overlooked conversations and evaluates the normative social views and discourse. Our approach models the stance of responses as a proxy for the underlying opinion, enabling a considerate and reflective representation of diverse social viewpoints. We evaluate the framework using both (i) positive-unlabeled (PU) online learning with base classifiers, and (ii) instruction-tuned language models to assess post-training alignment. Through this, we provide a principled and structured lens on how implicit opinions are (mis)represented and offer a pathway toward more inclusive model behavior.