In this study, we investigate the potential of language models (LMs) in aiding patients experiencing anomia, a difficulty identifying the names of items. Identifying the intended target item from patient's circumlocution involves the two challenges of term failure and error: (1) The terms relevant to identifying the item remain unseen. (2) What makes the challenge unique is inherent perturbed terms by semantic paraphasia, which are not exactly related to the target item, hindering the identification process. To address each, we propose robustifying the model from semantically paraphasic errors and enhancing the model with unseen terms with gradient-based selective augmentation. Specifically, the gradient value controls augmented data quality amid semantic errors, while the gradient variance guides the inclusion of unseen but relevant terms. Due to limited domain-specific datasets, we evaluate the model on the Tip-of-the-Tongue dataset as an intermediary task and then apply our findings to real patient data from AphasiaBank. Our results demonstrate strong performance against baselines, aiding anomia patients by addressing the outlined challenges.

Large language models (LLMs) exhibit remarkable reasoning capabilities across diverse downstream tasks. However, their autoregressive nature leads to substantial inference latency, posing challenges for real-time applications. Speculative sampling mitigates this issue by introducing a drafting phase followed by a parallel validation phase, enabling faster token generation and verification. Existing approaches, however, overlook the inherent coherence in text generation, limiting their efficiency. To address this gap, we propose a Speculative Sampling with Syntactic and Semantic Coherence (S$^4$C) framework, which extends speculative sampling by leveraging multi-head drafting for rapid token generation and a continuous verification tree for efficient candidate validation and feature reuse. Experimental results demonstrate that S$^4$C surpasses baseline methods across mainstream tasks, offering enhanced efficiency, parallelism, and the ability to generate more valid tokens with fewer computational resources. On Spec-bench benchmarks, S$^4$C achieves an acceleration ratio of 2.26x-2.60x, outperforming state-of-the-art methods.

Vertical-domain large language models (LLMs) play a crucial role in specialized scenarios such as finance, healthcare, and law; however, their training often relies on large-scale annotated data and substantial computational resources, impeding rapid development and continuous iteration. To address these challenges, we introduce the Collaborative Editable Model (CoEM), which constructs a candidate knowledge pool from user-contributed domain snippets, leverages interactive user-model dialogues combined with user ratings and attribution analysis to pinpoint high-value knowledge fragments, and injects these fragments via in-context prompts for lightweight domain adaptation. With high-value knowledge, the LLM can generate more accurate and domain-specific content. In a financial information scenario, we collect 15k feedback from about 120 users and validate CoEM with user ratings to assess the quality of generated insights, demonstrating significant improvements in domain-specific generation while avoiding the time and compute overhead of traditional fine-tuning workflows.

Accurate prediction of pedestrian trajectories is essential for applications in robotics and surveillance systems. While existing approaches primarily focus on social interactions between pedestrians, they often overlook the rich environmental context that significantly shapes human movement patterns. In this paper, we propose SceneAware, a novel framework that explicitly incorporates scene understanding to enhance trajectory prediction accuracy. Our method leverages a Vision Transformer~(ViT) scene encoder to process environmental context from static scene images, while Multi-modal Large Language Models~(MLLMs) generate binary walkability masks that distinguish between accessible and restricted areas during training. We combine a Transformer-based trajectory encoder with the ViT-based scene encoder, capturing both temporal dynamics and spatial constraints. The framework integrates collision penalty mechanisms that discourage predicted trajectories from violating physical boundaries, ensuring physically plausible predictions. SceneAware is implemented in both deterministic and stochastic variants. Comprehensive experiments on the ETH/UCY benchmark datasets show that our approach outperforms state-of-the-art methods, with more than 50\% improvement over previous models. Our analysis based on different trajectory categories shows that the model performs consistently well across various types of pedestrian movement. This highlights the importance of using explicit scene information and shows that our scene-aware approach is both effective and reliable in generating accurate and physically plausible predictions. Code is available at: https://github.com/juho127/SceneAware.

Modern deep learning is increasingly characterized by the use of open-weight foundation models that can be fine-tuned on specialized datasets. This has led to a proliferation of expert models and adapters, often shared via platforms like HuggingFace and AdapterHub. To leverage these resources, numerous model upcycling methods have emerged, enabling the reuse of fine-tuned models in multi-task systems. A natural pipeline has thus formed to harness the benefits of transfer learning and amortize sunk training costs: models are pre-trained on general data, fine-tuned on specific tasks, and then upcycled into more general-purpose systems. A prevailing assumption is that improvements at one stage of this pipeline propagate downstream, leading to gains at subsequent steps. In this work, we challenge that assumption by examining how expert fine-tuning affects model upcycling. We show that long fine-tuning of experts that optimizes for their individual performance leads to degraded merging performance, both for fully fine-tuned and LoRA-adapted models, and to worse downstream results when LoRA adapters are upcycled into MoE layers. We trace this degradation to the memorization of a small set of difficult examples that dominate late fine-tuning steps and are subsequently forgotten during merging. Finally, we demonstrate that a task-dependent aggressive early stopping strategy can significantly improve upcycling performance.

Yangliuqing woodblock prints, a cornerstone of China's intangible cultural heritage, are celebrated for their intricate designs and vibrant colors. However, preserving these traditional art forms while fostering innovation presents significant challenges. This study explores the DeepSeek + MidJourney approach to generating creative, themed Yangliuqing woodblock prints focused on the fight against COVID-19 and depicting joyous winners. Using Fréchet Inception Distance (FID) scores for evaluation, the method that combined DeepSeek-generated thematic prompts, MidJourney-generated thematic images, original Yangliuqing prints, and DeepSeek-generated key prompts in MidJourney-generated outputs achieved the lowest mean FID score (150.2) with minimal variability (σ = 4.9). Additionally, feedback from 62 participants, collected via questionnaires, confirmed that this hybrid approach produced the most representative results. Moreover, the questionnaire data revealed that participants demonstrated the highest willingness to promote traditional culture and the strongest interest in consuming the AI-generated images produced through this method. These findings underscore the effectiveness of an innovative approach that seamlessly blends traditional artistic elements with modern AI-driven creativity, ensuring both cultural preservation and contemporary relevance.

The Achilles heel of Large Language Models (LLMs) is hallucination, which has drastic consequences for the clinical domain. This is particularly important with regards to automatically generating discharge summaries (a lengthy medical document that summarizes a hospital in-patient visit). Automatically generating these summaries would free physicians to care for patients and reduce documentation burden. The goal of this work is to discover new methods that combine language-based graphs and deep learning models to address provenance of content and trustworthiness in automatic summarization. Our method shows impressive reliability results on the publicly available Medical Information Mart for Intensive III (MIMIC-III) corpus and clinical notes written by physicians at Anonymous Hospital. rovide our method, generated discharge ary output examples, source code and trained models.

--Retrieval-Augmented Generation (RAG) addresses limitations of large language models (LLMs) by leveraging a vector database to provide more accurate and up-to-date information. When a user submits a query, RAG executes a vector search to find relevant documents, which are then used to generate a response. However, ensuring the relevance of retrieved documents with a query would be a big challenge. T o address this, a secondary model, known as a relevant grader, can be served to verify its relevance. T o reduce computational requirements of a relevant grader, a lightweight small language model is preferred. Its precision is comparable to that of llama-3.1-70b. Large Language Models (LLMs) have achieved exceptional capabilities in various Natural Language Processing (NLP) tasks [1]-[3], demonstrating their ability to absorb and retain vast amounts of knowledge. When responding to specific queries, LLMs often provide informative answers, leveraging the extensive range of information they acquired during their training.

There is an unmet need to evaluate the language difficulty of short, conversational passages of text, particularly for training and filtering Large Language Models (LLMs). We introduce Ace-CEFR, a dataset of English conversational text passages expert-annotated with their corresponding level of text difficulty. We experiment with several models on Ace-CEFR, including Transformer-based models and LLMs. We show that models trained on Ace-CEFR can measure text difficulty more accurately than human experts and have latency appropriate to production environments. Finally, we release the Ace-CEFR dataset to the public for research and development.

Developing agents capable of exploring, planning and learning in complex open-ended environments is a grand challenge in artificial intelligence (AI). Hierarchical reinforcement learning (HRL) offers a promising solution to this challenge by discovering and exploiting the temporal structure within a stream of experience. The strong appeal of the HRL framework has led to a rich and diverse body of literature attempting to discover a useful structure. However, it is still not clear how one might define what constitutes good structure in the first place, or the kind of problems in which identifying it may be helpful. This work aims to identify the benefits of HRL from the perspective of the fundamental challenges in decision-making, as well as highlight its impact on the performance trade-offs of AI agents. Through these benefits, we then cover the families of methods that discover temporal structure in HRL, ranging from learning directly from online experience to offline datasets, to leveraging large language models (LLMs). Finally, we highlight the challenges of temporal structure discovery and the domains that are particularly well-suited for such endeavours.