Multi-Head Mixture-of-Experts (MH-MoE) demonstrates superior performance by using the multi-head mechanism to collectively attend to information from various representation spaces within different experts. In this paper, we present a novel implementation of MH-MoE that maintains both FLOPs and parameter parity with sparse Mixture of Experts models. Experimental results on language models show that the new implementation yields quality improvements over both vanilla MoE and fine-grained MoE models. Additionally, our experiments demonstrate that MH-MoE is compatible with 1-bit Large Language Models (LLMs) such as BitNet.

Image aesthetics is a crucial metric in the field of image generation. However, textual aesthetics has not been sufficiently explored. With the widespread application of large language models (LLMs), previous work has primarily focused on the correctness of content and the helpfulness of responses. Nonetheless, providing responses with textual aesthetics is also an important factor for LLMs, which can offer a cleaner layout and ensure greater consistency and coherence in content. Additionally, we develop two evaluation methods for textual aesthetics based on text and image analysis, respectively. Our experiments demonstrate that using textual aesthetics data and employing the TAPO fine-tuning method not only improves aesthetic scores but also enhances performance on general evaluation datasets such as AlpacalEval and Anera-hard. Image aesthetics (Huang et al., 2024a; Murray et al., 2012; Kong et al., 2016; Ke et al., 2021; Bosse et al., 2017) has emerged as a prominent research area within computer vision, focusing on assessing and improving the visual appeal of images. Aesthetics has recently been integrated into state-ofthe-art image generation models, such as diffusion models (Rombach et al., 2022), significantly enhancing the visual quality of generated images (Wu et al., 2024a; 2023) and aligning them more closely with human preferences (Huang et al., 2024a; Wu et al., 2024b; 2023). Meanwhile, advancements in large language models (LLMs) like ChatGPT (OpenAI, 2023) and LLaMA (Touvron et al., 2023b; Dubey et al., 2024) have demonstrated impressive generative capabilities across various domains, including code, articles, and web content. Although LLMs have made significant progress in generating textual content, enhancing the aesthetic quality of their output remains a critical challenge.

Deploying large language models (LLMs) is challenging due to their high memory and computational demands, especially during long-context inference. While keyvalue (KV) caching accelerates inference by reusing previously computed keys and values, it also introduces significant memory overhead. Existing KV cache compression methods--such as eviction and merging--typically compress the KV cache after it is generated and overlook the hidden states, failing to improve the speed of the prefilling stage. Additionally, applying a uniform compression rate across different attention heads can harm crucial retrieval heads in needle-in-ahaystack tasks due to excessive compression. In this paper, we propose UNComp, an uncertainty-aware compression scheme that leverages matrix entropy to estimate model uncertainty across layers and heads at the token sequence level. By grouping layers and heads based on their uncertainty, UNComp adaptively compresses both the hidden states and the KV cache. Our method achieves a 1.6 speedup in the prefilling stage and reduces the KV cache to 4.74% of its original size, resulting in a 6.4 increase in throughput and a 1.4 speedup in inference with only a 1.41% performance loss. Remarkably, in needle-in-a-haystack tasks, UNComp outperforms the full-size KV cache even when compressed to 9.38% of its original size. Our approach offers an efficient, training-free Grouped-Query Attention paradigm that can be seamlessly integrated into existing KV cache schemes. The proliferation of large language models (LLMs) has led to unprecedented advancements in natural language processing (Achiam et al., 2023; Kaplan et al., 2020), enabling capabilities ranging from simple text generation to complex reasoning and dialogue.

We present UncertaintyRAG, a novel approach for long-context Retrieval-Augmented Generation (RAG) that utilizes Signal-to-Noise Ratio (SNR)-based span uncertainty to estimate similarity between text chunks. This span uncertainty enhances model calibration, improving robustness and mitigating semantic inconsistencies introduced by random chunking. Leveraging this insight, we propose an efficient unsupervised learning technique to train the retrieval model, alongside an effective data sampling and scaling strategy. UncertaintyRAG outperforms baselines by 2.03% on LLaMA-2-7B, achieving state-of-the-art results while using only 4% of the training data compared to other advanced open-source retrieval models under distribution shift settings. Our method demonstrates strong calibration through span uncertainty, leading to improved generalization and robustness in long-context RAG tasks. Additionally, UncertaintyRAG provides a lightweight retrieval model that can be integrated into any large language model with varying context window lengths, without the need for fine-tuning, showcasing the flexibility of our approach.

Graph Neural Networks (GNNs) have become pivotal tools for a range of graph-based learning tasks. Notably, most current GNN architectures operate under the assumption of homophily, whether explicitly or implicitly. While this underlying assumption is frequently adopted, it is not universally applicable, which can result in potential shortcomings in learning effectiveness. In this paper, \textbf{for the first time}, we transfer the prevailing concept of ``one node one receptive field" to the heterophilic graph. By constructing a proxy label predictor, we enable each node to possess a latent prediction distribution, which assists connected nodes in determining whether they should aggregate their associated neighbors. Ultimately, every node can have its own unique aggregation hop and pattern, much like each snowflake is unique and possesses its own characteristics. Based on observations, we innovatively introduce the Heterophily Snowflake Hypothesis and provide an effective solution to guide and facilitate research on heterophilic graphs and beyond. We conduct comprehensive experiments including (1) main results on 10 graphs with varying heterophily ratios across 10 backbones; (2) scalability on various deep GNN backbones (SGC, JKNet, etc.) across various large number of layers (2,4,6,8,16,32 layers); (3) comparison with conventional snowflake hypothesis; (4) efficiency comparison with existing graph pruning algorithms. Our observations show that our framework acts as a versatile operator for diverse tasks. It can be integrated into various GNN frameworks, boosting performance in-depth and offering an explainable approach to choosing the optimal network depth. The source code is available at \url{https://github.com/bingreeky/HeteroSnoH}.

In the realm of autonomous driving, robust perception under out-of-distribution conditions is paramount for the safe deployment of vehicles. Challenges such as adverse weather, sensor malfunctions, and environmental unpredictability can severely impact the performance of autonomous systems. The 2024 RoboDrive Challenge was crafted to propel the development of driving perception technologies that can withstand and adapt to these real-world variabilities. Focusing on four pivotal tasks -- BEV detection, map segmentation, semantic occupancy prediction, and multi-view depth estimation -- the competition laid down a gauntlet to innovate and enhance system resilience against typical and atypical disturbances. This year's challenge consisted of five distinct tracks and attracted 140 registered teams from 93 institutes across 11 countries, resulting in nearly one thousand submissions evaluated through our servers. The competition culminated in 15 top-performing solutions, which introduced a range of innovative approaches including advanced data augmentation, multi-sensor fusion, self-supervised learning for error correction, and new algorithmic strategies to enhance sensor robustness. These contributions significantly advanced the state of the art, particularly in handling sensor inconsistencies and environmental variability. Participants, through collaborative efforts, pushed the boundaries of current technologies, showcasing their potential in real-world scenarios. Extensive evaluations and analyses provided insights into the effectiveness of these solutions, highlighting key trends and successful strategies for improving the resilience of driving perception systems. This challenge has set a new benchmark in the field, providing a rich repository of techniques expected to guide future research in this field.

Early and timely prediction of patient care demand not only affects effective resource allocation but also influences clinical decision-making as well as patient experience. Accurately predicting patient care demand, however, is a ubiquitous challenge for hospitals across the world due, in part, to the demand's time-varying temporal variability, and, in part, to the difficulty in modelling trends in advance. To address this issue, here, we develop two methods, a relatively simple time-vary linear model, and a more advanced neural network model. The former forecasts patient arrivals hourly over a week based on factors such as day of the week and previous 7-day arrival patterns. The latter leverages a long short-term memory (LSTM) model, capturing non-linear relationships between past data and a three-day forecasting window. We evaluate the predictive capabilities of the two proposed approaches compared to two na\"ive approaches - a reduced-rank vector autoregressive (VAR) model and the TBATS model. Using patient care demand data from Rambam Medical Center in Israel, our results show that both proposed models effectively capture hourly variations of patient demand. Additionally, the linear model is more explainable thanks to its simple architecture, whereas, by accurately modelling weekly seasonal trends, the LSTM model delivers lower prediction errors. Taken together, our explorations suggest the utility of machine learning in predicting time-varying patient care demand; additionally, it is possible to predict patient care demand with good accuracy (around 4 patients) three days or a week in advance using machine learning.

Sparse Mixtures of Experts (SMoE) scales model capacity without significant increases in training and inference costs, but exhibits the following two issues: (1) Low expert activation, where only a small subset of experts are activated for optimization. (2) Lacking fine-grained analytical capabilities for multiple semantic concepts within individual tokens. We propose Multi-Head Mixture-of-Experts (MH-MoE), which employs a multi-head mechanism to split each token into multiple sub-tokens. These sub-tokens are then assigned to and processed by a diverse set of experts in parallel, and seamlessly reintegrated into the original token form. The multi-head mechanism enables the model to collectively attend to information from various representation spaces within different experts, while significantly enhances expert activation, thus deepens context understanding and alleviate overfitting. Moreover, our MH-MoE is straightforward to implement and decouples from other SMoE optimization methods, making it easy to integrate with other SMoE models for enhanced performance. Extensive experimental results across three tasks: English-focused language modeling, Multi-lingual language modeling and Masked multi-modality modeling tasks, demonstrate the effectiveness of MH-MoE.

LoRA has gained widespread acceptance in the fine-tuning of large pre-trained models to cater to a diverse array of downstream tasks, showcasing notable effectiveness and efficiency, thereby solidifying its position as one of the most prevalent fine-tuning techniques. Due to the modular nature of LoRA's plug-and-play plugins, researchers have delved into the amalgamation of multiple LoRAs to empower models to excel across various downstream tasks. Nonetheless, extant approaches for LoRA fusion grapple with inherent challenges. Direct arithmetic merging may result in the loss of the original pre-trained model's generative capabilities or the distinct identity of LoRAs, thereby yielding suboptimal outcomes. On the other hand, Reference tuning-based fusion exhibits limitations concerning the requisite flexibility for the effective combination of multiple LoRAs. In response to these challenges, this paper introduces the Mixture of LoRA Experts (MoLE) approach, which harnesses hierarchical control and unfettered branch selection. The MoLE approach not only achieves superior LoRA fusion performance in comparison to direct arithmetic merging but also retains the crucial flexibility for combining LoRAs effectively. Extensive experimental evaluations conducted in both the Natural Language Processing (NLP) and Vision & Language (V&L) domains substantiate the efficacy of MoLE.