Personalizing AI systems requires understanding not just what users prefer, but the reasons that underlie those preferences - yet current preference models typically treat human judgment as a black box. We introduce PrefPalette, a framework that decomposes preferences into attribute dimensions and tailors its preference prediction to distinct social community values in a human-interpretable manner. PrefPalette operationalizes a cognitive science principle known as multi-attribute decision making in two ways: (1) a scalable counterfactual attribute synthesis step that involves generating synthetic training data to isolate for individual attribute effects (e.g., formality, humor, cultural values), and (2) attention-based preference modeling that learns how different social communities dynamically weight these attributes. This approach moves beyond aggregate preference modeling to capture the diverse evaluation frameworks that drive human judgment. When evaluated on 45 social communities from the online platform Reddit, PrefPalette outperforms GPT-4o by 46.6% in average prediction accuracy. Beyond raw predictive improvements, PrefPalette also shed light on intuitive, community-specific profiles: scholarly communities prioritize verbosity and stimulation, conflict-oriented communities value sarcasm and directness, and support-based communities emphasize empathy. By modeling the attribute-mediated structure of human judgment, PrefPalette delivers both superior preference modeling and transparent, interpretable insights, and serves as a first step toward more trustworthy, value-aware personalized applications.

Partner selection is crucial for cooperation and hinges on communication. As artificial agents, especially those powered by large language models (LLMs), become more autonomous, intelligent, and persuasive, they compete with humans for partnerships. Yet little is known about how humans select between human and AI partners and adapt under AI-induced competition pressure. We constructed a communication-based partner selection game and examined the dynamics in hybrid mini-societies of humans and bots powered by a state-of-the-art LLM. Through three experiments (N = 975), we found that bots, though more prosocial than humans and linguistically distinguishable, were not selected preferentially when their identity was hidden. Instead, humans misattributed bots' behaviour to humans and vice versa. Disclosing bots' identity induced a dual effect: it reduced bots' initial chances of being selected but allowed them to gradually outcompete humans by facilitating human learning about the behaviour of each partner type. These findings show how AI can reshape social interaction in mixed societies and inform the design of more effective and cooperative hybrid systems.

--Large Language Models (LLMs) offer significant promise for intelligent traffic management; however, current chain-based systems like TrafficGPT are hindered by sequential task execution, high token usage, and poor scalability, making them inefficient for complex, real-world scenarios. T o address these limitations, we propose GraphTrafficGPT, a novel graph-based architecture, which fundamentally redesigns the task coordination process for LLM-driven traffic applications. Graph-TrafficGPT represents tasks and their dependencies as nodes and edges in a directed graph, enabling efficient parallel execution and dynamic resource allocation. The main idea behind the proposed model is a Brain Agent that decomposes user queries, constructs optimized dependency graphs, and coordinates a network of specialized agents for data retrieval, analysis, visualization, and simulation. By introducing advanced context-aware token management and supporting concurrent multi-query processing, the proposed architecture handles interdependent tasks typical of modern urban mobility environments. Experimental results demonstrate that GraphTrafficGPT reduces token consumption by 50.2% and average response latency by 19.0% compared to TrafficGPT, while supporting simultaneous multi-query execution with up to 23.0% improvement in efficiency. Large Language Models (LLMs) have changed artificial intelligence capabilities across domains by enabling natural language understanding and generation at new levels. The recent models, such as GPT -4, Claude, and Llama, can comprehend complex instructions, reason through problems, and generate coherent responses across diverse applications [1].

Aim. There are 10s of thousands of code review comments each week at Meta. We developed Metamate for Code Review (MetaMateCR) that provides AI-assisted fixes for reviewer comments in production at scale. Method. We developed an internal benchmark of 64k data points to fine-tune Llama models. Once our models achieve reasonable offline results, we roll them into production. To ensure that our AI-assisted fixes do not negatively impact the time it takes to do code reviews, we conduct randomized controlled safety trials as well as full production experiments. Offline Results. As a baseline, we compare GPT-4o to our small and large Llama models. In offline results, our LargeLSFT model creates an exact match patch 68% of the time outperforming GPT-4o by 9 percentage points (pp). The internal models also use more modern Hack functions when compared to the PHP functions suggested by GPT-4o. Safety Trial. When we roll MetaMateCR into production in a safety trial that compares no AI patches with AI patch suggestions, we see a large regression with reviewers taking over 5% longer to conduct reviews. After investigation, we modify the UX to only show authors the AI patches, and see no regressions in the time for reviews. Production. When we roll LargeLSFT into production, we see an ActionableToApplied rate of 19.7%, which is a 9.2pp improvement over GPT-4o. Our results illustrate the importance of safety trials in ensuring that AI does not inadvertently slow down engineers, and a successful review comment to AI patch product running at scale.

There has been extensive research on assessing the value orientation of Large Language Models (LLMs) as it can shape user experiences across demographic groups. However, several challenges remain. First, while the Multiple Choice Question (MCQ) setting has been shown to be vulnerable to perturbations, there is no systematic comparison of probing methods for value probing. Second, it is unclear to what extent the probed values capture in-context information and reflect models' preferences for real-world actions. In this paper, we evaluate the robustness and expressiveness of value representations across three widely used probing strategies. We use variations in prompts and options, showing that all methods exhibit large variances under input perturbations. We also introduce two tasks studying whether the values are responsive to demographic context, and how well they align with the models' behaviors in value-related scenarios. We show that the demographic context has little effect on the free-text generation, and the models' values only weakly correlate with their preference for value-based actions. Our work highlights the need for a more careful examination of LLM value probing and awareness of its limitations.

The safety of large language models (LLMs) has garnered significant research attention. In this paper, we argue that previous empirical studies demonstrate LLMs exhibit a propensity to trust information from authoritative sources, such as academic papers, implying new possible vulnerabilities. To verify this possibility, a preliminary analysis is designed to illustrate our two findings. Based on this insight, a novel jailbreaking method, Paper Summary Attack (\llmname{PSA}), is proposed. It systematically synthesizes content from either attack-focused or defense-focused LLM safety paper to construct an adversarial prompt template, while strategically infilling harmful query as adversarial payloads within predefined subsections. Extensive experiments show significant vulnerabilities not only in base LLMs, but also in state-of-the-art reasoning model like Deepseek-R1. PSA achieves a 97\% attack success rate (ASR) on well-aligned models like Claude3.5-Sonnet and an even higher 98\% ASR on Deepseek-R1. More intriguingly, our work has further revealed diametrically opposed vulnerability bias across different base models, and even between different versions of the same model, when exposed to either attack-focused or defense-focused papers. This phenomenon potentially indicates future research clues for both adversarial methodologies and safety alignment.Code is available at https://github.com/233liang/Paper-Summary-Attack

Video generation models have achieved remarkable progress in creating high-quality, photorealistic content. However, their ability to accurately simulate physical phenomena remains a critical and unresolved challenge. This paper presents PhyWorldBench, a comprehensive benchmark designed to evaluate video generation models based on their adherence to the laws of physics. The benchmark covers multiple levels of physical phenomena, ranging from fundamental principles like object motion and energy conservation to more complex scenarios involving rigid body interactions and human or animal motion. Additionally, we introduce a novel ""Anti-Physics"" category, where prompts intentionally violate real-world physics, enabling the assessment of whether models can follow such instructions while maintaining logical consistency. Besides large-scale human evaluation, we also design a simple yet effective method that could utilize current MLLM to evaluate the physics realism in a zero-shot fashion. We evaluate 12 state-of-the-art text-to-video generation models, including five open-source and five proprietary models, with a detailed comparison and analysis. we identify pivotal challenges models face in adhering to real-world physics. Through systematic testing of their outputs across 1,050 curated prompts-spanning fundamental, composite, and anti-physics scenarios-we identify pivotal challenges these models face in adhering to real-world physics. We then rigorously examine their performance on diverse physical phenomena with varying prompt types, deriving targeted recommendations for crafting prompts that enhance fidelity to physical principles.

AutoML has advanced in handling complex tasks using the integration of LLMs, yet its efficiency remains limited by dependence on specific underlying tools. In this paper, we introduce LightAutoDS-Tab, a multi-AutoML agentic system for tasks with tabular data, which combines an LLM-based code generation with several AutoML tools. Our approach improves the flexibility and robustness of pipeline design, outperforming state-of-the-art open-source solutions on several data science tasks from Kaggle. The code of LightAutoDS-Tab is available in the open repository https://github.com/sb-ai-lab/LADS

--Large language models like GPT -4, Gemini, and Claude have transformed natural language processing (NLP) tasks such as question answering, dialogue generation, summarization, and so forth; yet their susceptibility to hallucination stands as one of the major challenges. Among numerous approaches to overcome this challenge, integration of Knowledge Graphs (KGs) into language models has emerged as a promising solution as it provides structured, reliable, domain-specific, and up-to-date external information to the language models. In this paper, we introduce ALIGNed-LLM, a simple yet effective approach to improve language models' factuality via a lean strategy to infuse KGs into the latent space of language models inspired by LLaV A where visual and textual information is infused. We use embeddings from a pre-trained Knowledge Graph Embedding (KGE) model, such as TransE, and a trainable projection layer to align entity and text embeddings. This alignment enables the language model to distinguish between similar entities improving factual grounding and reducing hallucination. We tested our approach on three popular questions-answering benchmark datasets alongside language models of varying sizes, showing significant improvement. Furthermore, we applied our approach to a real-world financial use case from a large central bank in Europe, which demands high accuracy and precision, demonstrating a substantial improvement of the LLM answers. The emergence of Large Language Models (LLMs) such as GPT -4 [27], Gemini [1], Llama [36], and Claude [2], is producing a performance revolution across natural language processing tasks, such as question answering, dialogue generation, summarization, and many more. LLMs have also confirmed the universality of human language, showing their potential to build domain-specific assistants, trained to follow natural language instructions and accomplish various tasks end to end [22]. The opinions expressed in this paper are personal and should not be attributed to Banca d'Italia. The work was done outside Amazon.

Recently, many benchmarks and datasets have been developed to evaluate Vision-Language Models (VLMs) using visual question answering (VQA) pairs, and models have shown significant accuracy improvements. However, these benchmarks rarely test the model's ability to accurately complete visual entailment, for instance, accepting or refuting a hypothesis based on the image. To address this, we propose COREVQA (Crowd Observations and Reasoning Entailment), a benchmark of 5608 image and synthetically generated true/false statement pairs, with images derived from the CrowdHuman dataset, to provoke visual entailment reasoning on challenging crowded images. Our results show that even the top-performing VLMs achieve accuracy below 80%, with other models performing substantially worse (39.98%-69.95%). This significant performance gap reveals key limitations in VLMs' ability to reason over certain types of image-question pairs in crowded scenes.