Human Activity Recognition (HAR), which uses data from Inertial Measurement Unit (IMU) sensors, has many practical applications in healthcare and assisted living environments. However, its use in real-world scenarios has been limited by the lack of comprehensive IMU-based HAR datasets that cover a wide range of activities and the lack of transparency in existing HAR models. Zero-shot HAR (ZS-HAR) overcomes the data limitations, but current models struggle to explain their decisions, making them less transparent. This paper introduces a novel IMU-based ZS-HAR model called the Self-Explainable Zero-shot Human Activity Recognition Network (SEZ-HARN). It can recognize activities not encountered during training and provide skeleton videos to explain its decision-making process. We evaluate the effectiveness of the proposed SEZ-HARN on four benchmark datasets PAMAP2, DaLiAc, HTD-MHAD and MHealth and compare its performance against three state-of-the-art black-box ZS-HAR models. The experiment results demonstrate that SEZ-HARN produces realistic and understandable explanations while achieving competitive Zero-shot recognition accuracy. SEZ-HARN achieves a Zero-shot prediction accuracy within 3\% of the best-performing black-box model on PAMAP2 while maintaining comparable performance on the other three datasets.

Post-training processes are essential phases in grounding pre-trained language models to real-world tasks, with learning from demonstrations or preference signals playing a crucial role in this adaptation. We present a unified theoretical framework bridging Supervised Fine-Tuning (SFT) and preference learning in Large Language Model (LLM) post-training. Through rigorous mathematical derivation, we demonstrate that both SFT and preference learning methods like Direct Preference Optimization (DPO) operate within the same optimal policy-reward subspace, with SFT representing a special case of implicit reward learning. Our analysis reveals a critical limitation in conventional SFT: the KL divergence term in distribution matching becomes constant with respect to the policy during optimization, failing to constrain model updates. To address this, we propose a simple yet effective learning rate reduction approach that yields significant performance improvements (up to \textbf{25\%} relative gain and \textbf{6\%} absolute win rate increase in instruction following tasks. Additionally, we derive alternative SFT objectives from various f-divergence functions that preserve the KL term during optimization, further enhancing post-DPO model performance. Finally, we extend the theoretical relationship between LLM logits and Q-functions from preference learning to the SFT context, providing mathematical derivations and experimental validation.

Multimodal large language models (MLLMs) have shown impressive capabilities across tasks involving both visual and textual modalities. However, growing concerns remain about their potential to encode and amplify gender bias, particularly in socially sensitive applications. Existing benchmarks predominantly evaluate bias in isolated scenarios, overlooking how bias may emerge subtly through interpersonal interactions. We fill this gap by going beyond single-entity evaluation and instead focusing on a deeper examination of relational and contextual gender bias in dual-individual interactions. We introduce Genres, a novel benchmark designed to evaluate gender bias in MLLMs through the lens of social relationships in generated narratives. Genres assesses gender bias through a dual-character profile and narrative generation task that captures rich interpersonal dynamics and supports a fine-grained bias evaluation suite across multiple dimensions. Experiments on both open- and closed-source MLLMs reveal persistent, context-sensitive gender biases that are not evident in single-character settings. Our findings underscore the importance of relationship-aware benchmarks for diagnosing subtle, interaction-driven gender bias in MLLMs and provide actionable insights for future bias mitigation.

Large language models are thought to have potential to aid in medical decision making. We investigate this here. We start with the treatment problem, the patient's core medical decision-making task, which is solved in collaboration with a healthcare provider. We discuss approaches to solving the treatment problem, including -- within evidence-based medicine -- trials and observational data. We then discuss the chat problem, and how this differs from the treatment problem -- in particular as it relates to imitation. We then discuss how a large language model might be used to solve the treatment problem and highlight some of the challenges that emerge. We finally discuss how these challenges relate to evidence-based medicine, and how this might inform next steps.

Large language models (LLMs) have demonstrated considerable reasoning abilities in various tasks such as mathematics and coding. However, recent studies indicate that even the best models lack true comprehension of their reasoning processes. In this paper, we explore how self-play can enhance the rationality of models in the reasoning process without supervision from humans or superior models. We design a Critic-Discernment Game(CDG) in which a prover first provides a solution to a given problem and is subsequently challenged by critiques of its solution. These critiques either aim to assist or mislead the prover. The objective of the prover is to maintain the correct answer when faced with misleading comments, while correcting errors in response to constructive feedback. Our experiments on tasks involving mathematical reasoning, stepwise error detection, self-correction, and long-chain reasoning demonstrate that CDG training can significantly improve the ability of well-aligned LLMs to comprehend their reasoning process.

Generative AI and large language models (LLMs) offer significant potential for automating the extraction of structured information from web pages. In this work, we focus on food product pages from online retailers and explore schema-constrained extraction approaches to retrieve key product attributes, such as ingredient lists and nutrition tables. We compare two LLM-based approaches, direct extraction and indirect extraction via generated functions, evaluating them in terms of accuracy, efficiency, and cost on a curated dataset of 3,000 food product pages from three different online shops. Our results show that although the indirect approach achieves slightly lower accuracy (96.48\%, $-1.61\%$ compared to direct extraction), it reduces the number of required LLM calls by 95.82\%, leading to substantial efficiency gains and lower operational costs. These findings suggest that indirect extraction approaches can provide scalable and cost-effective solutions for large-scale information extraction tasks from template-based web pages using LLMs.

Large Reasoning Models (LRMs), which autonomously produce a reasoning Chain of Thought (CoT) before producing final responses, offer a promising approach to interpreting and monitoring model behaviors. Inspired by the observation that certain CoT patterns -- e.g., ``Wait, did I miss anything?'' -- consistently emerge across tasks, we explore whether LRMs exhibit human-like cognitive habits. Building on Habits of Mind, a well-established framework of cognitive habits associated with successful human problem-solving, we introduce CogTest, a principled benchmark designed to evaluate LRMs' cognitive habits. CogTest includes 16 cognitive habits, each instantiated with 25 diverse tasks, and employs an evidence-first extraction method to ensure reliable habit identification. With CogTest, we conduct a comprehensive evaluation of 16 widely used LLMs (13 LRMs and 3 non-reasoning ones). Our findings reveal that LRMs, unlike conventional LLMs, not only exhibit human-like habits but also adaptively deploy them according to different tasks. Finer-grained analyses further uncover patterns of similarity and difference in LRMs' cognitive habit profiles, particularly certain inter-family similarity (e.g., Qwen-3 models and DeepSeek-R1). Extending the study to safety-related tasks, we observe that certain habits, such as Taking Responsible Risks, are strongly associated with the generation of harmful responses. These findings suggest that studying persistent behavioral patterns in LRMs' CoTs is a valuable step toward deeper understanding of LLM misbehavior. The code is available at: https://github.com/jianshuod/CogTest.

In the realm of contemporary social media, automatic stance detection is pivotal for opinion mining, as it synthesizes and examines user perspectives on contentious topics to uncover prevailing trends and sentiments. Traditional stance detection research often targets individual instances, thereby limiting its capacity to model multi-party discussions typical in real social media scenarios. This shortcoming largely stems from the scarcity of datasets that authentically capture the dynamics of social media interactions, hindering advancements in conversational stance detection. In this paper, we introduce MT2-CSD, a comprehensive dataset for multi-target, multi-turn conversational stance detection. To the best of our knowledge, MT2-CSD is the largest dataset available for this purpose, comprising 24,457 annotated instances and exhibiting the greatest conversational depth, thereby presenting new challenges for stance detection. To address these challenges, we propose the Large Language model enhanced Conversational Relational Attention Network (LLM-CRAN), which exploits the reasoning capabilities of LLMs to improve conversational understanding. We conduct extensive experiments to evaluate the efficacy of LLM-CRAN on the MT2-CSD dataset. The experimental results indicate that LLM-CRAN significantly outperforms strong baseline models in the task of conversational stance detection.

Graph-based Retrieval-Augmented Generation (Graph-RAG) enhances large language models (LLMs) by structuring retrieval over an external corpus. However, existing approaches typically assume a static corpus, requiring expensive full-graph reconstruction whenever new documents arrive, limiting their scalability in dynamic, evolving environments. To address these limitations, we introduce EraRAG, a novel multi-layered Graph-RAG framework that supports efficient and scalable dynamic updates. Our method leverages hyperplane-based Locality-Sensitive Hashing (LSH) to partition and organize the original corpus into hierarchical graph structures, enabling efficient and localized insertions of new data without disrupting the existing topology. The design eliminates the need for retraining or costly recomputation while preserving high retrieval accuracy and low latency. Experiments on large-scale benchmarks demonstrate that EraRag achieves up to an order of magnitude reduction in update time and token consumption compared to existing Graph-RAG systems, while providing superior accuracy performance. This work offers a practical path forward for RAG systems that must operate over continually growing corpora, bridging the gap between retrieval efficiency and adaptability. Our code and data are available at https://github.com/EverM0re/EraRAG-Official.

AI agents have experienced a paradigm shift, from early dominance by reinforcement learning (RL) to the rise of agents powered by large language models (LLMs), and now further advancing towards a synergistic fusion of RL and LLM capabilities. This progression has endowed AI agents with increasingly strong abilities. Despite these advances, to accomplish complex real-world tasks, agents are required to plan and execute effectively, maintain reliable memory, and coordinate smoothly with other agents. Achieving these capabilities involves contending with ever-present intricate information, operations, and interactions. In light of this challenge, data structurization can play a promising role by transforming intricate and disorganized data into well-structured forms that agents can more effectively understand and process. In this context, graphs, with their natural advantage in organizing, managing, and harnessing intricate data relationships, present a powerful data paradigm for structurization to support the capabilities demanded by advanced AI agents. To this end, this survey presents a first systematic review of how graphs can empower AI agents. Specifically, we explore the integration of graph techniques with core agent functionalities, highlight notable applications, and identify prospective avenues for future research. By comprehensively surveying this burgeoning intersection, we hope to inspire the development of next-generation AI agents equipped to tackle increasingly sophisticated challenges with graphs. Related resources are collected and continuously updated for the community in the Github link.