Recent large language model (LLM)-driven chat assistant systems have integrated memory components to track user-assistant chat histories, enabling more accurate and personalized responses. However, their long-term memory capabilities in sustained interactions remain underexplored. This paper introduces LongMemEval, a comprehensive benchmark designed to evaluate five core long-term memory abilities of chat assistants: information extraction, multi-session reasoning, temporal reasoning, knowledge updates, and abstention. With 500 meticulously curated questions embedded within freely scalable user-assistant chat histories, LongMemEval presents a significant challenge to existing long-term memory systems, with commercial chat assistants and long-context LLMs showing 30% accuracy drop on memorizing information across sustained interactions. We then present a unified framework that breaks down the long-term memory design into four design choices across the indexing, retrieval, and reading stages. Built upon key experimental insights, we propose several memory designs including session decomposition for optimizing value granularity, fact-augmented key expansion for enhancing the index structure, and time-aware query expansion for refining the search scope. Experiment results show that these optimizations greatly improve both memory recall and downstream question answering on LongMemEval. Overall, our study provides valuable resources and guidance for advancing the long-term memory capabilities of LLM-based chat assistants, paving the way toward more personalized and reliable conversational AI.

Alzheimer's Disease (AD) is a prevalent neurodegenerative condition where early detection is vital. Handwriting, often affected early in AD, offers a non-invasive and cost-effective way to capture subtle motor changes. State-of-the-art research on handwriting, mostly online, based AD detection has predominantly relied on manually extracted features, fed as input to shallow machine learning models. Some recent works have proposed deep learning (DL)-based models, either 1D-CNN or 2D-CNN architectures, with performance comparing favorably to handcrafted schemes. These approaches, however, overlook the intrinsic relationship between the 2D spatial patterns of handwriting strokes and their 1D dynamic characteristics, thus limiting their capacity to capture the multimodal nature of handwriting data. Moreover, the application of Transformer models remains basically unexplored. To address these limitations, we propose a novel approach for AD detection, consisting of a learnable multimodal hybrid attention model that integrates simultaneously 2D handwriting images with 1D dynamic handwriting signals. Our model leverages a gated mechanism to combine similarity and difference attention, blending the two modalities and learning robust features by incorporating information at different scales. Our model achieved state-of-the-art performance on the DARWIN dataset, with an F1-score of 90.32\% and accuracy of 90.91\% in Task 8 ('L' writing), surpassing the previous best by 4.61% and 6.06% respectively.

Figure 1: HART is an early autoregressive model that can directly generate 1024 1024 images with quality comparable to diffusion models, while offering significantly improved efficiency. It achieves 4.5-7.7 higher throughput, 3.1-5.9 Check out our online demo and video. We introduce Hybrid Autoregressive Transformer (HART), an autoregressive (AR) visual generation model capable of directly generating 1024 1024 images, rivaling diffusion models in image generation quality. Existing AR models face limitations due to the poor image reconstruction quality of their discrete tokenizers and the prohibitive training costs associated with generating 1024px images. To address these challenges, we present the hybrid tokenizer, which decomposes the continuous latents from the autoencoder into two components: discrete tokens representing the big picture and continuous tokens representing the residual components that cannot be represented by the discrete tokens. The discrete component is modeled by a scalable-resolution discrete AR model, while the continuous component is learned with a lightweight residual diffusion module with only 37M parameters. Compared with the discrete-only VAR tokenizer, our hybrid approach improves reconstruction FID from 2.11 to 0.30 on MJHQ-30K, leading to a 31% generation FID improvement from 7.85 to 5.38. HART also outperforms state-of-the-art diffusion models in both FID and CLIP score, with 4.5-7.7 higher throughput and 6.9-13.4 Part of the work was done when Haotian Tang and Shang Yang were summer interns at NVIDIA. Prompt: A panda that has been cybernetically enhanced.

Target-speaker speech processing (TS) tasks, such as target-speaker automatic speech recognition (TS-ASR), target speech extraction (TSE), and personal voice activity detection (p-VAD), are important for extracting information about a desired speaker's speech even when it is corrupted by interfering speakers. While most studies have focused on training schemes or system architectures for each specific task, the auxiliary network for embedding target-speaker cues has not been investigated comprehensively in a unified cross-task evaluation. Therefore, this paper aims to address a fundamental question: what is the preferred speaker embedding for TS tasks? To this end, for the TS-ASR, TSE, and p-VAD tasks, we compare pre-trained speaker encoders (i.e., self-supervised or speaker recognition models) that compute speaker embeddings from pre-recorded enrollment speech of the target speaker with ideal speaker embeddings derived directly from the target speaker's identity in the form of a one-hot vector. To further understand the properties of ideal speaker embedding, we optimize it using a gradient-based approach to improve performance on the TS task. Our analysis reveals that speaker verification performance is somewhat unrelated to TS task performances, the one-hot vector outperforms enrollment-based ones, and the optimal embedding depends on the input mixture.

Large Language Models (LLMs) have demonstrated impressive performance on a wide range of natural language processing (NLP) tasks, primarily through in-context learning (ICL). In ICL, the LLM is provided with examples that represent a given task such that it learns to generate answers for test inputs. However, access to these in-context examples is not guaranteed especially for low-resource or massively multilingual tasks. In this work, we propose an unsupervised approach to mine in-context examples for machine translation (MT), enabling unsupervised MT (UMT) across different languages. Our approach begins with word-level mining to acquire word translations that are then used to perform sentence-level mining. As the quality of mined parallel pairs may not be optimal due to noise or mistakes, we introduce a filtering criterion to select the optimal in-context examples from a pool of unsupervised parallel sentences. We evaluate our approach using two multilingual LLMs on 288 directions from the FLORES-200 dataset and analyze the impact of various linguistic features on performance. Our findings demonstrate the effectiveness of our unsupervised approach in mining in-context examples for MT, leading to better or comparable translation performance as translation with regular in-context samples (extracted from human-annotated data), while also outperforming the other state-of-the-art UMT methods by an average of $7$ BLEU points.

Background: Conventional prediction methods such as logistic regression and gradient boosting have been widely utilized for disease onset prediction for their reliability and interpretability. Deep learning methods promise enhanced prediction performance by extracting complex patterns from clinical data, but face challenges like data sparsity and high dimensionality. Methods: This study compares conventional and deep learning approaches to predict lung cancer, dementia, and bipolar disorder using observational data from eleven databases from North America, Europe, and Asia. Models were developed using logistic regression, gradient boosting, ResNet, and Transformer, and validated both internally and externally across the data sources. Discrimination performance was assessed using AUROC, and calibration was evaluated using Eavg. Findings: Across 11 datasets, conventional methods generally outperformed deep learning methods in terms of discrimination performance, particularly during external validation, highlighting their better transportability. Learning curves suggest that deep learning models require substantially larger datasets to reach the same performance levels as conventional methods. Calibration performance was also better for conventional methods, with ResNet showing the poorest calibration. Interpretation: Despite the potential of deep learning models to capture complex patterns in structured observational healthcare data, conventional models remain highly competitive for disease onset prediction, especially in scenarios involving smaller datasets and if lengthy training times need to be avoided. The study underscores the need for future research focused on optimizing deep learning models to handle the sparsity, high dimensionality, and heterogeneity inherent in healthcare datasets, and find new strategies to exploit the full capabilities of deep learning methods.

Characters are at the heart of every story, driving the plot and engaging readers. In this study, we explore the understanding of characters in full-length books, which contain complex narratives and numerous interacting characters. We define two tasks: character description, which generates a brief factual profile, and character analysis, which offers an in-depth interpretation, including character development, personality, and social context. We introduce the BookWorm dataset, pairing books from the Gutenberg Project with human-written descriptions and analyses. Using this dataset, we evaluate state-of-the-art long-context models in zero-shot and fine-tuning settings, utilizing both retrieval-based and hierarchical processing for book-length inputs. Our findings show that retrieval-based approaches outperform hierarchical ones in both tasks. Additionally, fine-tuned models using coreference-based retrieval produce the most factual descriptions, as measured by fact- and entailment-based metrics. We hope our dataset, experiments, and analysis will inspire further research in character-based narrative understanding.

The term "K-12" stands for "Kindergarten through 12th grade" and represents the full range of primary and secondary education. Within this system, a strong emphasis has been placed on STEM (Science, Technology, Engineering, and Mathematics) education as a means to prepare students for a technology-driven future. STEM education at the K-12 level focuses on building foundational knowledge in scientific inquiry, technological literacy, engineering principles, and mathematical reasoning [10, 29, 64]. The K-12 STEM education emphasizes interdisciplinary learning, where students apply concepts from multiple domains to solve real-world challenges, such as integrating mathematics with science to tackle engineering problems [29]. The importance of K-12 STEM education lies in its ability to prepare students for a rapidly evolving, technology-driven world by fostering critical thinking, creativity, and problem-solving skills from an early age [10]. Students who engage in well-structured STEM curricula are more likely to pursue further education and careers in high-demand fields like information technology and engineering which are essential for technological innovation. Additionally, K-12 STEM education equips students with competencies such as analytical thinking, which prepare them for a wide range of career paths while enabling them to tackle complex problems [64]. Recognizing the importance of STEM education at the K-12 level, it is essential to deliver K-12 STEM education at scale to ensure equitable access to individual students.

Interpretability is a topic that has been in the spotlight for the past few years. Most existing interpretability techniques produce interpretations in the form of rules or feature importance. These interpretations, while informative, may be harder to understand for non-expert users and therefore, cannot always be considered as adequate explanations. To that end, explanations in natural language are often preferred, as they are easier to comprehend and also more presentable to end-users. This work introduces an early concept for a novel pipeline that can be used in text classification tasks, offering predictions and explanations in natural language. It comprises of two models: a classifier for labelling the text and an explanation generator which provides the explanation. The proposed pipeline can be adopted by any text classification task, given that ground truth rationales are available to train the explanation generator. Our experiments are centred around the tasks of sentiment analysis and offensive language identification in Greek tweets, using a Greek Large Language Model (LLM) to obtain the necessary explanations that can act as rationales. The experimental evaluation was performed through a user study based on three different metrics and achieved promising results for both datasets.

An ideal speech recognition model has the capability to transcribe speech accurately under various characteristics of speech signals, such as speaking style (read and spontaneous), speech context (formal and informal), and background noise conditions (clean and moderate). Building such a model requires a significant amount of training data with diverse speech characteristics. Currently, Indonesian data is dominated by read, formal, and clean speech, leading to a scarcity of Indonesian data with other speech variabilities. To develop Indonesian automatic speech recognition (ASR), we present our research on state-of-the-art speech recognition models, namely Massively Multilingual Speech (MMS) and Whisper, as well as compiling a dataset comprising Indonesian speech with variabilities to facilitate our study. We further investigate the models' predictive ability to transcribe Indonesian speech data across different variability groups. The best results were achieved by the Whisper fine-tuned model across datasets with various characteristics, as indicated by the decrease in word error rate (WER) and character error rate (CER). Moreover, we found that speaking style variability affected model performance the most.