The paper introduces a 12,000-instance dataset of two-option moral dilemmas covering three pairwise value conflicts—Honesty vs. Justice, Justice vs. Autonomy, and Autonomy vs. Honesty—translated into Hindi, Arabic, Spanish, and Chinese to test cross‑lingual behavior. Benchmarking on GPT‑5‑mini shows a consistent preference for Honesty over Autonomy across all languages when no policy is provided, while Llama‑3.2‑1/3B models exhibit a strong first‑option bias that is largely eliminated by plain fine‑tuning or Direct Preference Optimization, raising accuracy above 98%. The authors propose a task vector transfer method that orthogonalizes value preference vectors with respect to general instruction‑following vectors, effectively isolating specific value directions and enabling task arithmetic to flip a model’s stance.
arXiv:2601.10307v2 Announce Type: replace Abstract: Enhancing the moral alignment of Large Language Models (LLMs) is a critical challenge in AI safety. Current alignment techniques often act as super...
The paper argues that aligning large language models (LLMs) at the level of latent representations—specifically by matching their internal categorization of moral concepts to human prototype-based judgments—improves safety. Current alignment methods that focus on observable responses fail to preserve fine-grained moral categorization, leaving models vulnerable to adversarial rephrasings. By optimizing representational similarity, the authors demonstrate that LLMs can maintain more robust moral categorization and exhibit better adversarial robustness across multiple benchmarks and model sizes.
arXiv:2608. 14629v1 Announce Type: cross Abstract: As Large Language Models (LLMs) become the mainstay for information retrieval and summarization tasks, ensuring that they are always non-partisan and invulnerable to political bias is a critical step towards safer and more trustworthy Artificial Intelligence (AI).
arXiv:2609.08515v1 Announce Type: cross Abstract: As Large Language Models (LLMs) are increasingly integrated into human society, aligning them with pluralistic social values has become a critical pr...
arXiv:2603.16017v2 Announce Type: replace-cross Abstract: Large language models (LLMs) increasingly participate in morally sensitive decision-making, yet how they organize ethical frameworks across r...
The paper introduces a multilingual story moral generation task to evaluate cultural alignment in large language models. Using a dataset of human-written story morals from 14 language‑culture pairs, the authors compare model outputs to human interpretations through semantic similarity, a preference survey, and value categorization. They find that advanced models like GPT‑4o and Gemini produce morally similar and preferred responses but show less cross‑linguistic variation, focusing on a narrower set of shared values, indicating a limitation in capturing the diversity of human narrative understanding.
The paper argues that AI alignment depends on a system’s ability to exhibit a coherent moral policy—stable, monotonic, decisive, and Pareto‑viable—rather than on any specific moral standard. The authors test nine large language models across varied moral scenarios and find that none maintain consistent verdicts, with surface‑form changes causing up to 99% shifts in outcomes. This indicates that current LLM agents lack the structural moral competence required for meaningful alignment.
arXiv:2605. 22660v2 Announce Type: replace-cross Abstract: Moral language is subtle and culturally variable, making it difficult to translate faithfully across languages.
The paper introduces CoT-Interpretability Alignment (CIA), a metric that quantifies how well a large language model’s chain-of-thought (CoT) explanations match its internal reasoning processes. Evaluated on two-hop question answering, hint intervention, and integer multiplication across three LLMs, the study finds limited alignment (44.8–75.9%) and demonstrates that post‑training with a reward combining task accuracy and parametric faithfulness can substantially improve CoT faithfulness without sacrificing accuracy. The authors provide a framework for auditing CoT faithfulness and a pathway to making explicit reasoning more trustworthy, with code and data publicly available.
arXiv:2603. 00048v2 Announce Type: replace-cross Abstract: Large Language Models (LLMs) are increasingly deployed in sensitive applications including psychological support, healthcare, and high-stakes decision-making.
MBBQ (Multilingual Bias Benchmark for Question-answering) is a curated extension of the English BBQ dataset that includes Dutch, Spanish, and Turkish, designed to measure stereotypes across these languages while controlling for cultural differences and task accuracy. The dataset also provides a parallel control set to assess question-answering performance independently of bias. Experiments with several open-source and proprietary LLMs show that non‑English languages can exhibit greater bias than English, and cross‑lingual bias differences persist except in the most accurate models.
arXiv:2607. 16232v1 Announce Type: cross Abstract: The growing use of statistical learning algorithms to infer human preferences from high-dimensional choice data runs up against a fundamental challenge: choice alternatives typically differ in many ways simultaneously, so it is generally unclear which factors actually drove an observed decision and should be credited as preferences.