Recent advances in multimodal large language models (MLLMs) have significantly improved medical AI, enabling it to unify the understanding of visual and textual information. However, as medical knowledge continues to evolve, it is critical to allow these models to efficiently update outdated or incorrect information without retraining from scratch. Although textual knowledge editing has been widely studied, there is still a lack of systematic benchmarks for multi-modal medical knowledge editing involving image and text modalities. To fill this gap, we present MedMKEB, the first comprehensive benchmark designed to evaluate the reliability, generality, locality, portability, and robustness of knowledge editing in medical multimodal large language models. MedMKEB is built on a high-quality medical visual question-answering dataset and enriched with carefully constructed editing tasks, including counterfactual correction, semantic generalization, knowledge transfer, and adversarial robustness. We incorporate human expert validation to ensure the accuracy and reliability of the benchmark. Extensive single editing and sequential editing experiments on state-of-the-art general and medical MLLMs demonstrate the limitations of existing knowledge-based editing approaches in medicine, highlighting the need to develop specialized editing strategies. MedMKEB will serve as a standard benchmark to promote the development of trustworthy and efficient medical knowledge editing algorithms.

Web navigation represents a critical and challenging domain for evaluating artificial general intelligence (AGI), demanding complex decision-making within high-entropy, dynamic environments with combinatorially explosive action spaces. Current approaches to building autonomous web agents either focus on offline imitation learning or online exploration, but rarely integrate both paradigms effectively. Inspired by the dual-process theory of human cognition, we derive a principled decomposition into fast System 1 and slow System 2 cognitive processes. This decomposition provides a unifying perspective on existing web agent methodologies, bridging the gap between offline learning of intuitive reactive behaviors and online acquisition of deliberative planning capabilities. We implement this framework in CogniWeb, a modular agent architecture that adaptively toggles between fast intuitive processing and deliberate reasoning based on task complexity. Our evaluation on WebArena demonstrates that CogniWeb achieves competitive performance (43.96% success rate) while maintaining significantly higher efficiency (75% reduction in token usage).

Parameter-Efficient Fine-Tuning (PEFT) is essential for adapting Large Language Models (LLMs). In practice, LLMs are often required to handle a diverse set of tasks from multiple domains, a scenario naturally addressed by multi-task learning (MTL). Within this MTL context, a prevailing trend involves LoRA variants with multiple adapters or heads, which advocate for structural diversity to capture task-specific knowledge. Our findings present a direct challenge to this paradigm. We first show that a simplified multi-head architecture with high inter-head similarity substantially outperforms complex multi-adapter and multi-head systems. This leads us to question the multi-component paradigm itself, and we further demonstrate that a standard single-adapter LoRA, with a sufficiently increased rank, also achieves highly competitive performance. These results lead us to a new hypothesis: effective MTL generalization hinges on learning robust shared representations, not isolating task-specific features. To validate this, we propose Align-LoRA, which incorporates an explicit loss to align task representations within the shared adapter space. Experiments confirm that Align-LoRA significantly surpasses all baselines, establishing a simpler yet more effective paradigm for adapting LLMs to multiple tasks.

Dialogues Aspect-based Sentiment Quadruple Extraction (DiaASQ) aims to extract all target-aspect-opinion-sentiment quadruples from a given multi-round, multi-participant dialogue. Existing methods typically learn word relations across entire dialogues, assuming a uniform distribution of sentiment elements. However, we find that dialogues often contain multiple semantically independent sub-dialogues without clear dependencies between them. Therefore, learning word relationships across the entire dialogue inevitably introduces additional noise into the extraction process. To address this, our method focuses on partitioning dialogues into semantically independent sub-dialogues. Achieving completeness while minimizing these sub-dialogues presents a significant challenge. Simply partitioning based on reply relationships is ineffective. Instead, we propose utilizing a structural entropy minimization algorithm to partition the dialogues. This approach aims to preserve relevant utterances while distinguishing irrelevant ones as much as possible. Furthermore, we introduce a two-step framework for quadruple extraction: first extracting individual sentiment elements at the utterance level, then matching quadruples at the sub-dialogue level. Extensive experiments demonstrate that our approach achieves state-of-the-art performance in DiaASQ with much lower computational costs.

Visual navigation in unknown environments based solely on natural language descriptions is a key capability for intelligent robots. In this work, we propose a navigation framework built upon off-the-shelf Visual Language Models (VLMs), enhanced with two human-inspired mechanisms: perspective-based active grounding, which dynamically adjusts the robot's viewpoint for improved visual inspection, and historical memory backtracking, which enables the system to retain and re-evaluate uncertain observations over time. Unlike existing approaches that passively rely on incidental visual inputs, our method actively optimizes perception and leverages memory to resolve ambiguity, significantly improving vision-language grounding in complex, unseen environments. Our framework operates in a zero-shot manner, achieving strong generalization to diverse and open-ended language descriptions without requiring labeled data or model fine-tuning. Experimental results on Habitat-Matterport 3D (HM3D) show that our method outperforms state-of-the-art approaches in language-driven object navigation. We further demonstrate its practicality through real-world deployment on a quadruped robot, achieving robust and effective navigation performance.

Large language models (LLMs) have shown strong reasoning capabilities when fine-tuned with reinforcement learning (RL). However, such methods require extensive data and compute, making them impractical for smaller models. Current approaches to curriculum learning or data selection are largely heuristic-driven or demand extensive computational resources, limiting their scalability and generalizability. We propose \textbf{SPaRFT}, a self-paced learning framework that enables efficient learning based on the capability of the model being trained through optimizing which data to use and when. First, we apply \emph{cluster-based data reduction} to partition training data by semantics and difficulty, extracting a compact yet diverse subset that reduces redundancy. Then, a \emph{multi-armed bandit} treats data clusters as arms, optimized to allocate training samples based on model current performance. Experiments across multiple reasoning benchmarks show that SPaRFT achieves comparable or better accuracy than state-of-the-art baselines while using up to \(100\times\) fewer samples. Ablation studies and analyses further highlight the importance of both data clustering and adaptive selection. Our results demonstrate that carefully curated, performance-driven training curricula can unlock strong reasoning abilities in LLMs with minimal resources.

Modern LLM pipelines increasingly resemble data-centric systems: they retrieve external context, compose intermediate outputs, validate results, and adapt based on runtime feedback. Yet, the central element guiding this process -- the prompt -- remains a brittle, opaque string, disconnected from the surrounding dataflow. This disconnect limits reuse, optimization, and runtime control. In this paper, we describe our vision and an initial design for SPEAR, a language and runtime that fills this prompt management gap by making prompts structured, adaptive, and first-class components of the execution model. SPEAR enables (1) runtime prompt refinement -- modifying prompts dynamically in response to execution-time signals such as confidence, latency, or missing context; and (2) structured prompt management -- organizing prompt fragments into versioned views with support for introspection and logging. SPEAR defines a prompt algebra that governs how prompts are constructed and adapted within a pipeline. It supports multiple refinement modes (manual, assisted, and automatic), giving developers a balance between control and automation. By treating prompt logic as structured data, SPEAR enables optimizations such as operator fusion, prefix caching, and view reuse. Preliminary experiments quantify the behavior of different refinement modes compared to static prompts and agentic retries, as well as the impact of prompt-level optimizations such as operator fusion.

Background: Understanding social determinants of health (SDoH) factors contributing to suicide incidents is crucial for early intervention and prevention. However, data-driven approaches to this goal face challenges such as long-tailed factor distributions, analyzing pivotal stressors preceding suicide incidents, and limited model explainability. Methods: We present a multi-stage large language model framework to enhance SDoH factor extraction from unstructured text. Our approach was compared to other state-of-the-art language models (i.e., pre-trained BioBERT and GPT-3.5-turbo) and reasoning models (i.e., DeepSeek-R1). We also evaluated how the model's explanations help people annotate SDoH factors more quickly and accurately. The analysis included both automated comparisons and a pilot user study. Results: We show that our proposed framework demonstrated performance boosts in the overarching task of extracting SDoH factors and in the finer-grained tasks of retrieving relevant context. Additionally, we show that fine-tuning a smaller, task-specific model achieves comparable or better performance with reduced inference costs. The multi-stage design not only enhances extraction but also provides intermediate explanations, improving model explainability. Conclusions: Our approach improves both the accuracy and transparency of extracting suicide-related SDoH from unstructured texts. These advancements have the potential to support early identification of individuals at risk and inform more effective prevention strategies.

This paper introduces a comprehensive benchmark for evaluating how Large Language Models (LLMs) respond to linguistic shibboleths: subtle linguistic markers that can inadvertently reveal demographic attributes such as gender, social class, or regional background. Through carefully constructed interview simulations using 100 validated question-response pairs, we demonstrate how LLMs systematically penalize certain linguistic patterns, particularly hedging language, despite equivalent content quality. Our benchmark generates controlled linguistic variations that isolate specific phenomena while maintaining semantic equivalence, which enables the precise measurement of demographic bias in automated evaluation systems. We validate our approach along multiple linguistic dimensions, showing that hedged responses receive 25.6% lower ratings on average, and demonstrate the benchmark's effectiveness in identifying model-specific biases. This work establishes a foundational framework for detecting and measuring linguistic discrimination in AI systems, with broad applications to fairness in automated decision-making contexts.

Multimodal Large Language Models (MLLMs) demonstrate remarkable capabilities but often struggle with complex, multi-step mathematical reasoning, where minor errors in visual perception or logical deduction can lead to complete failure. While Process Reward Models (PRMs) offer step-by-step supervision, existing multimodal PRMs are limited to being binary verifiers that can identify but not correct errors, offering little explanatory power. To address these deficiencies, we introduce the Generative Multimodal Process Reward Model (GM-PRM), a novel paradigm that transforms the PRM from a passive judge into an active reasoning collaborator. Instead of a simple scalar score, GM-PRM provides a fine-grained, interpretable analysis of each reasoning step, evaluating its step intent, visual alignment, and logical soundness. More critically, GM-PRM is trained to generate a corrected version of the first erroneous step it identifies. This unique corrective capability enables our new test-time inference strategy, Refined Best-of-N (Refined-BoN). This framework actively enhances solution quality by using the PRM's generated correction to guide the policy model toward a more promising reasoning trajectory, thereby improving the diversity and correctness of the solution pool. We demonstrate that GM-PRM achieves state-of-the-art results on multiple multimodal math benchmarks, significantly boosting policy model performance with remarkable data efficiency, requiring only a 20K-sample training dataset. Our code will be released upon acceptance.