Cross-modal entity linking refers to the ability to align entities and their attributes across different modalities. While cross-modal entity linking is a fundamental skill needed for real-world applications such as multimodal code generation, fake news detection, or scene understanding, it has not been thoroughly studied in the literature. In this paper, we introduce a new task and benchmark to address this gap. Our benchmark, MATE, consists of 5.5k evaluation instances featuring visual scenes aligned with their textual representations. To evaluate cross-modal entity linking performance, we design a question-answering task that involves retrieving one attribute of an object in one modality based on a unique attribute of that object in another modality. We evaluate state-of-the-art Vision-Language Models (VLMs) and humans on this task, and find that VLMs struggle significantly compared to humans, particularly as the number of objects in the scene increases. Our analysis also shows that, while chain-of-thought prompting can improve VLM performance, models remain far from achieving human-level proficiency. These findings highlight the need for further research in cross-modal entity linking and show that MATE is a strong benchmark to support that progress.

The application of Large Language Models (LLMs) to various clinical applications has attracted growing research attention. However, real-world clinical decision-making differs significantly from the standardized, exam-style scenarios commonly used in current efforts. In this paper, we present the RiskAgent system to perform a broad range of medical risk predictions, covering over 387 risk scenarios across diverse complex diseases, e.g., cardiovascular disease and cancer. RiskAgent is designed to collaborate with hundreds of clinical decision tools, i.e., risk calculators and scoring systems that are supported by evidence-based medicine. To evaluate our method, we have built the first benchmark MedRisk specialized for risk prediction, including 12,352 questions spanning 154 diseases, 86 symptoms, 50 specialties, and 24 organ systems. The results show that our RiskAgent, with 8 billion model parameters, achieves 76.33% accuracy, outperforming the most recent commercial LLMs, o1, o3-mini, and GPT-4.5, and doubling the 38.39% accuracy of GPT-4o. On rare diseases, e.g., Idiopathic Pulmonary Fibrosis (IPF), RiskAgent outperforms o1 and GPT-4.5 by 27.27% and 45.46% accuracy, respectively. Finally, we further conduct a generalization evaluation on an external evidence-based diagnosis benchmark and show that our RiskAgent achieves the best results. These encouraging results demonstrate the great potential of our solution for diverse diagnosis domains. To improve the adaptability of our model in different scenarios, we have built and open-sourced a family of models ranging from 1 billion to 70 billion parameters. Our code, data, and models are all available at https://github.com/AI-in-Health/RiskAgent.

This work examines the integration of large language models (LLMs) into multi-agent simulations by replacing the hard-coded programs of agents with LLM-driven prompts. The proposed approach is showcased in the context of two examples of complex systems from the field of swarm intelligence: ant colony foraging and bird flocking. Central to this study is a toolchain that integrates LLMs with the NetLogo simulation platform, leveraging its Python extension to enable communication with GPT-4o via the OpenAI API. This toolchain facilitates prompt-driven behavior generation, allowing agents to respond adaptively to environmental data. For both example applications mentioned above, we employ both structured, rule-based prompts and autonomous, knowledge-driven prompts. Our work demonstrates how this toolchain enables LLMs to study self-organizing processes and induce emergent behaviors within multi-agent environments, paving the way for new approaches to exploring intelligent systems and modeling swarm intelligence inspired by natural phenomena. We provide the code, including simulation files and data at https://github.com/crjimene/swarm_gpt.

Large Language Models have demonstrated outstanding performance across various downstream tasks and have been widely applied in multiple scenarios. Human-annotated preference data is used for training to further improve LLMs' performance, which is constrained by the upper limit of human performance. Therefore, Self-Rewarding method has been proposed, where LLMs generate training data by rewarding their own outputs. However, the existing self-rewarding paradigm is not effective in mathematical reasoning scenarios and may even lead to a decline in performance. In this work, we propose the Process-based Self-Rewarding pipeline for language models, which introduces long-thought reasoning, step-wise LLM-as-a-Judge, and step-wise preference optimization within the self-rewarding paradigm. Our new paradigm successfully enhances the performance of LLMs on multiple mathematical reasoning benchmarks through iterative Process-based Self-Rewarding, demonstrating the immense potential of self-rewarding to achieve LLM reasoning that may surpass human capabilities.

In this paper, we investigate how model distillation impacts the development of reasoning features in large language models (LLMs). To explore this, we train a crosscoder on Qwen-series models and their fine-tuned variants. Our results suggest that the crosscoder learns features corresponding to various types of reasoning, including self-reflection and computation verification. Moreover, we observe that distilled models contain unique reasoning feature directions, which could be used to steer the model into over-thinking or incisive-thinking mode. In particular, we perform analysis on four specific reasoning categories: (a) self-reflection, (b) deductive reasoning, (c) alternative reasoning, and (d) contrastive reasoning. Finally, we examine the changes in feature geometry resulting from the distillation process and find indications that larger distilled models may develop more structured representations, which correlate with enhanced distillation performance. By providing insights into how distillation modifies the model, our study contributes to enhancing the transparency and reliability of AI systems.

Existing training-time safety alignment techniques for large language models (LLMs) remain vulnerable to jailbreak attacks. Direct preference optimization (DPO), a widely deployed alignment method, exhibits limitations in both experimental and theoretical contexts as its loss function proves suboptimal for refusal learning. Through gradient-based analysis, we identify these shortcomings and propose an improved safety alignment that disentangles DPO objectives into two components: (1) robust refusal training, which encourages refusal even when partial unsafe generations are produced, and (2) targeted unlearning of harmful knowledge. This approach significantly increases LLM robustness against a wide range of jailbreak attacks, including prefilling, suffix, and multi-turn attacks across both in-distribution and out-of-distribution scenarios. Furthermore, we introduce a method to emphasize critical refusal tokens by incorporating a reward-based token-level weighting mechanism for refusal learning, which further improves the robustness against adversarial exploits. Our research also suggests that robustness to jailbreak attacks is correlated with token distribution shifts in the training process and internal representations of refusal and harmful tokens, offering valuable directions for future research in LLM safety alignment. The code is available at https://github.com/wicai24/DOOR-Alignment

Large language models (LLMs) are trained on enormous documents that contain extensive world knowledge. However, it is still not well-understood how knowledge is acquired via autoregressive pre-training. This lack of understanding greatly hinders effective knowledge learning, especially for continued pretraining on up-to-date information, as this evolving information often lacks diverse repetitions like foundational knowledge. In this paper, we focus on understanding and improving LLM knowledge learning. We found and verified that knowledge learning for LLMs can be deemed as an implicit supervised task hidden in the autoregressive pre-training objective. Our findings suggest that knowledge learning for LLMs would benefit from methods designed to improve generalization ability for supervised tasks. Based on our analysis, we propose the formatting-based data augmentation to grow in-distribution samples, which does not present the risk of altering the facts embedded in documents as text paraphrasing. We also introduce sharpness-aware minimization as an effective optimization algorithm to better improve generalization. Moreover, our analysis and method can be readily extended to instruction tuning. Extensive experiment results validate our findings and demonstrate our methods' effectiveness in both continued pre-training and instruction tuning. This paper offers new perspectives and insights to interpret and design effective strategies for LLM knowledge learning.

High-quality data resources play a crucial role in learning large language models (LLMs), particularly for low-resource languages like Cantonese. Despite having more than 85 million native speakers, Cantonese is still considered a low-resource language in the field of natural language processing (NLP) due to factors such as the dominance of Mandarin, lack of cohesion within the Cantonese-speaking community, diversity in character encoding and input methods, and the tendency of overseas Cantonese speakers to prefer using English. In addition, rich colloquial vocabulary of Cantonese, English loanwords, and code-switching characteristics add to the complexity of corpus collection and processing. To address these challenges, we collect Cantonese texts from a variety of sources, including open source corpora, Hong Kong-specific forums, Wikipedia, and Common Crawl data. We conduct rigorous data processing through language filtering, quality filtering, content filtering, and de-duplication steps, successfully constructing a high-quality Cantonese corpus of over 2 billion tokens for training large language models. We further refined the model through supervised fine-tuning (SFT) on curated Cantonese tasks, enhancing its ability to handle specific applications. Upon completion of the training, the model achieves state-of-the-art (SOTA) performance on four Cantonese benchmarks. After training on our dataset, the model also exhibits improved performance on other mainstream language tasks.

LLM-based multi-agent systems (MAS) have shown significant potential in tackling diverse tasks. However, to design effective MAS, existing approaches heavily rely on manual configurations or multiple calls of advanced LLMs, resulting in inadaptability and high inference costs. In this paper, we simplify the process of building an MAS by reframing it as a generative language task, where the input is a user query and the output is a corresponding MAS. To address this novel task, we unify the representation of MAS as executable code and propose a consistency-oriented data construction pipeline to create a high-quality dataset comprising coherent and consistent query-MAS pairs. Using this dataset, we train MAS-GPT, an open-source medium-sized LLM that is capable of generating query-adaptive MAS within a single LLM inference. The generated MAS can be seamlessly applied to process user queries and deliver high-quality responses. Extensive experiments on 9 benchmarks and 5 LLMs show that the proposed MAS-GPT consistently outperforms 10+ baseline MAS methods on diverse settings, indicating MAS-GPT's high effectiveness, efficiency and strong generalization ability. Code will be available at https://github.com/rui-ye/MAS-GPT.

We present Attentive Reasoning Queries (ARQs), a novel structured reasoning approach that significantly improves instruction-following in Large Language Models through domain-specialized reasoning blueprints. While LLMs demonstrate remarkable capabilities across diverse tasks, they often fail to maintain adherence to complex, use-case-specific instructions during multi-turn conversations, presenting challenges for business-critical applications. ARQs address this limitation by guiding LLMs through systematic reasoning steps with targeted queries that reinstate critical instructions and facilitate intermediate reasoning throughout the completion process. In extensive testing within Parlant, our framework for reliable customer-facing agents in which ARQs were born out of necessity, they achieved a 90.2% success rate across 87 test scenarios, outperforming both Chain-of-Thought reasoning (86.1%) and direct response generation (81.5%). ARQs showed particular strength in addressing persistent failure modes like guideline re-application and hallucination prevention. Our analysis also revealed that ARQs can potentially be more computationally efficient than free-form reasoning when carefully designed. These findings demonstrate that structured reasoning approaches provide effective mechanisms for controlling how LLMs process information and make decisions in complex scenarios.