arXiv:2604.03244v1 Announce Type: new Abstract: AI evaluations have become the primary evidence for deploying generative AI systems across high-stakes domains. However, current evaluation paradigms often exhibit systemic validity failures. These issues, ranging from unjustified design choices to misaligned metrics, remain intractable without a principled framework for gathering validity evidence and conducting granular diagnostic analysis. In this position paper, we argue that item-level AI benchmark data is essential for establishing a rigorous science of AI evaluation. Item-level analysis enables fine-grained diagnostics and principled validation of benchmarks. We substantiate this position by dissecting current validity failures and revisiting evaluation paradigms across computer science and psychometrics. Through illustrative analyses of item properties and latent constructs, we demonstrate the unique insights afforded by item-level data. To catalyze community-wide adoption, we introduce OpenEval, a growing repository of item-level benchmark data designed supporting evidence-centered AI evaluation.

Nature, Published online: 01 April 2026; doi:10.1038/s41586-026-10303-2

A fully automated methodology based on rubrics capturing a broad range of cognitive and intellectual demands is illustrated using LLMs and tasks, demonstrating a new way to evaluate the capabilities of AI systems and anticipate their performance.

arXiv:2602.10625v3 Announce Type: replace Abstract: Theory of Mind (ToM) assesses whether models can infer hidden mental states such as beliefs, desires, and intentions, which is essential for natural social interaction. Although recent progress in Large Reasoning Models (LRMs) has boosted step-by-step inference in mathematics and coding, it is still underexplored whether this benefit transfers to socio-cognitive skills. We present a systematic study of nine advanced Large Language Models (LLMs), comparing reasoning models with non-reasoning models on three representative ToM benchmarks. The results show that reasoning models do not consistently outperform non-reasoning models and sometimes perform worse. A fine-grained analysis reveals three insights. First, slow thinking collapses: accuracy significantly drops as responses grow longer, and larger reasoning budgets hurt performance. Second, moderate and adaptive reasoning benefits performance: constraining reasoning length mitigates failure, while distinct success patterns demonstrate the necessity of dynamic adaptation. Third, option matching shortcut: when multiple choice options are removed, reasoning models improve markedly, indicating reliance on option matching rather than genuine deduction. We also design two intervention approaches: Slow-to-Fast (S2F) adaptive reasoning and Think-to-Match (T2M) shortcut prevention to further verify and mitigate the problems. With all results, our study highlights the advancement of LRMs in formal reasoning (e.g., math, code) cannot be fully transferred to ToM, a typical task in social reasoning. We conclude that achieving robust ToM requires developing unique capabilities beyond existing reasoning methods.

arXiv:2603.02983v1 Announce Type: cross Abstract: LLM agents increasingly act on users' personal information, yet existing privacy defenses remain limited in both design and adaptability. Most prior approaches rely on static or passive defenses, such as prompting and guarding. These paradigms are insufficient for supporting contextual, proactive privacy decisions in multi-step agent execution. We propose Contextualized Defense Instructing (CDI), a new privacy defense paradigm in which an instructor model generates step-specific, context-aware privacy guidance during execution, proactively shaping actions rather than merely constraining or vetoing them. Crucially, CDI is paired with an experience-driven optimization framework that trains the instructor via reinforcement learning (RL), where we convert failure trajectories with privacy violations into learning environments. We formalize baseline defenses and CDI as distinct intervention points in a canonical agent loop, and compare their privacy-helpfulness trade-offs within a unified simulation framework. Results show that our CDI consistently achieves a better balance between privacy preservation (94.2%) and helpfulness (80.6%) than baselines, with superior robustness to adversarial conditions and generalization.