Personalized value alignment has become increasingly important as large language models (LLMs) are expected to accommodate diverse user preferences. However, existing methods typically align model outputs with a static value profile across prompts, overlooking that the salience of value dimensions varies substantially across contexts. Inspired by Lewin’s Field Theory, which views human behavior as jointly shaped by personal dispositions and situational constraints, we model personal values as priors and context-dependent preferences as posteriors. We propose BaCVA, an inference-time Bayesian Context-aware personalized Value Alignment method that approximates posterior personalized preferences by integrating static personal values with scenario-specific value salience. BaCVA first estimates contextual value salience from generally normative responses, and then employs a dual-view personalization module to infer posterior preferences from complementary personal-value and scenario-driven perspectives. This Bayesian formulation enables more accurate and adaptive personalized value alignment while improving data efficiency via prior values. Extensive experiments on benchmarks demonstrate its superiority over strong baselines.

As LLMs evolve from standalone assistants into autonomous agents, they increasingly operate in populations where communication, coordination, and competition form community-like interaction networks. In such settings, individually value-steered behaviors can accumulate into group-level dynamics, raising the question of how human values shape collective outcomes in interacting LLM-agent populations. We introduce CIVA, a community-level value analysis framework that varies value direction and prevalence, then measures how these interventions propagate through repeated interactions into macro-level dynamics and micro-level behaviors. We apply it in a resource-constrained multi-agent community where LLM agents autonomously communicate, cooperate, and compete. The value scan reveals distinct structural response patterns: activation-supporting values such as benevolence supports persistence and resilience, an activation-constraining value such as tradition regulates growth and cohesion, and a bidirectionally disruptive value such as power perturbs collective regimes in both directions. At the micro level, representative interventions induce emergent behaviors including deception, betrayal, self-preservation, and power-seeking. These results show that human values can act as population-level structural variables in agent communities and motivate evaluation for multi-agent value alignment.

Generative retrieval offers a unified paradigm for personalized search by autoregressively decoding items as Semantic IDs. However, personalized search hinges on reasoning about query-history relevance, in which explicit textual reasoning incurs heavy cost and high latency, whereas purely implicit reasoning does not provide a structured intermediate latent interface for retrieval. To resolve this tension, we propose LaPR, an adaptive latent retrieval reasoning framework that turns explicit retrieval reasoning into compact latent signals. Specifically, LaPR constructs verified step-wise retrieval traces, induces step-conditioned semantic slots via vector quantization, and distills textual reasoning into these slots through a text-latent to pure-latent curriculum. At inference, LaPR adaptively predicts the reasoning depth and corresponding slots from the query and user history, providing latent reasoning signals for efficient Semantic-ID generation. Extensive experiments show that LaPR consistently surpasses all baselines, achieving SOTA personal retrieval performance while substantially reducing reasoning-token cost and inference latency.