Mixture of Hidden-Dimensions Transformer

Transformer models encounter challenges in scaling hidden dimensions efficiently, as uniformly increasing them inflates computational and memory costs while failing to emphasize the most relevant features for each token. For further understanding, we study hidden dimension sparsity and observe that trained Transformers utilize only a small fraction of token dimensions, revealing an "activation flow" pattern. Notably, there are shared sub-dimensions with sustained activation across multiple consecutive tokens and specialized sub-dimensions uniquely activated for each token. To better model token-relevant sub-dimensions, we propose MoHD (Mixture of Hidden Dimensions), a sparse conditional activation architecture. Particularly, MoHD employs shared sub-dimensions for common token features and a routing mechanism to dynamically activate specialized sub-dimensions. To mitigate potential information loss from sparsity, we design activation scaling and group fusion mechanisms to preserve activation flow. In this way, MoHD expands hidden dimensions with negligible increases in computation or parameters, efficient training and inference while maintaining performance. Evaluations across 10 NLP tasks show that MoHD surpasses Vanilla Transformers in parameter efficiency and task performance. It achieves 1.7% higher performance with 50% fewer activation parameters and 3.7% higher performance with a 3x parameter expansion at constant activation cost. MOHD offers a new perspective for scaling the model, showcasing the potential of hidden dimension sparsity to boost efficiency

On-Road Object Importance Estimation: A New Dataset and A Model with Multi-Fold Top-Down Guidance

This paper addresses the problem of on-road object importance estimation, which utilizes video sequences captured from the driver's perspective as the input. Although this problem is significant for safer and smarter driving systems, the exploration of this problem remains limited. On one hand, publicly-available large-scale datasets are scarce in the community. To address this dilemma, this paper contributes a new large-scale dataset named Traffic Object Importance (TOI). On the other hand, existing methods often only consider either bottom-up feature or single-fold guidance, leading to limitations in handling highly dynamic and diverse traffic scenarios. Different from existing methods, this paper proposes a model that integrates multi-fold top-down guidance with the bottom-up feature. Specifically, three kinds of top-down guidance factors (ie, driver intention, semantic context, and traffic rule) are integrated into our model. These factors are important for object importance estimation, but none of the existing methods simultaneously consider them. To our knowledge, this paper proposes the first on-road object importance estimation model that fuses multi-fold top-down guidance factors with bottom-up feature. Extensive experiments demonstrate that our model outperforms state-of-the-art methods by large margins, achieving 23.1% Average Precision (AP) improvement compared with the recently proposed model (ie, Goal).

MoR: Mixture of Ranks for Low-Rank Adaptation Tuning

Low-Rank Adaptation (LoRA) drives research to align its performance with full fine-tuning. However, significant challenges remain: (1) Simply increasing the rank size of LoRA does not effectively capture high-rank information, which leads to a performance bottleneck.(2) MoE-style LoRA methods substantially increase parameters and inference latency, contradicting the goals of efficient fine-tuning and ease of application. To address these challenges, we introduce Mixture of Ranks (MoR), which learns rank-specific information for different tasks based on input and efficiently integrates multi-rank information. We firstly propose a new framework that equates the integration of multiple LoRAs to expanding the rank of LoRA. Moreover, we hypothesize that low-rank LoRA already captures sufficient intrinsic information, and MoR can derive high-rank information through mathematical transformations of the low-rank components. Thus, MoR can reduces the learning difficulty of LoRA and enhances its multi-task capabilities. MoR achieves impressive results, with MoR delivering a 1.31\% performance improvement while using only 93.93\% of the parameters compared to baseline methods.

Exploiting Moving Arrays for Near-Field Sensing

This letter exploits moving arrays to enable nearfield multiple-input multiple-output (MIMO) sensing via a limited number of antenna elements. We consider a scenario where a base station (BS) is equipped with a uniform linear array (ULA) on a moving platform. The objective is to locate a point target in the two-dimensional (2D) space by leveraging the near-field channel characteristics created by the movement of antenna arrays. Under this setup, we analyze the Cramer-Rao bound (CRB) for estimating the target's 2D coordinate, which provides the fundamental sensing performance limits for localization. It is revealed that our proposed design with a moving array achieves a CRB that is proportional to the CRB obtained by an equivalent extremely large ULA matching the platform's size. This shows that the movement of antenna array significantly enlarges its effective aperture to enable near-field sensing. Numerical results show that the proposed moving array design substantially enhances the target estimation performance compared to the conventional fixed array benchmark.

Unleashing the Power of Emojis in Texts via Self-supervised Graph Pre-Training

Emojis have gained immense popularity on social platforms, serving as a common means to supplement or replace text. However, existing data mining approaches generally either completely ignore or simply treat emojis as ordinary Unicode characters, which may limit the model's ability to grasp the rich semantic information in emojis and the interaction between emojis and texts. Thus, it is necessary to release the emoji's power in social media data mining. To this end, we first construct a heterogeneous graph consisting of three types of nodes, i.e. post, word and emoji nodes to improve the representation of different elements in posts. The edges are also well-defined to model how these three elements interact with each other. To facilitate the sharing of information among post, word and emoji nodes, we propose a graph pre-train framework for text and emoji co-modeling, which contains two graph pre-training tasks: node-level graph contrastive learning and edge-level link reconstruction learning. Extensive experiments on the Xiaohongshu and Twitter datasets with two types of downstream tasks demonstrate that our approach proves significant improvement over previous strong baseline methods.

Non-target Divergence Hypothesis: Toward Understanding Domain Gaps in Cross-Modal Knowledge Distillation

Compared to single-modal knowledge distillation, cross-modal knowledge distillation faces more severe challenges due to domain gaps between modalities. Although various methods have proposed various solutions to overcome these challenges, there is still limited research on how domain gaps affect cross-modal knowledge distillation. This paper provides an in-depth analysis and evaluation of this issue. We first introduce the Non-Target Divergence Hypothesis (NTDH) to reveal the impact of domain gaps on cross-modal knowledge distillation. Our key finding is that domain gaps between modalities lead to distribution differences in non-target classes, and the smaller these differences, the better the performance of cross-modal knowledge distillation. Subsequently, based on Vapnik-Chervonenkis (VC) theory, we derive the upper and lower bounds of the approximation error for cross-modal knowledge distillation, thereby theoretically validating the NTDH. Finally, experiments on five cross-modal datasets further confirm the validity, generalisability, and applicability of the NTDH.

NACL: A General and Effective KV Cache Eviction Framework for LLMs at Inference Time

Large Language Models (LLMs) have ignited an innovative surge of AI applications, marking a new era of exciting possibilities equipped with extended context windows. However, hosting these models is cost-prohibitive mainly due to the extensive memory consumption of KV Cache involving long-context modeling. Despite several works proposing to evict unnecessary tokens from the KV Cache, most of them rely on the biased local statistics of accumulated attention scores and report performance using unconvincing metric like perplexity on inadequate short-text evaluation. In this paper, we propose NACL, a general framework for long-context KV cache eviction that achieves more optimal and efficient eviction in a single operation during the encoding phase. Due to NACL's efficiency, we combine more accurate attention score statistics in PROXY TOKENS EVICTION with the diversified random eviction strategy of RANDOM EVICTION, aiming to alleviate the issue of attention bias and enhance the robustness in maintaining pivotal tokens for long-context modeling tasks. Notably, our method significantly improves the performance on short- and long-text tasks by 80% and 76% respectively, reducing KV Cache by up to 50% with over 95% performance maintenance. The code is available at https: //github.com/PaddlePaddle/Research/ tree/master/NLP/ACL2024-NACL.

DHA: Learning Decoupled-Head Attention from Transformer Checkpoints via Adaptive Heads Fusion

Large language models (LLMs) with billions of parameters demonstrate impressive performance. However, the widely used Multi-Head Attention (MHA) in LLMs incurs substantial computational and memory costs during inference. While some efforts have optimized attention mechanisms by pruning heads or sharing parameters among heads, these methods often lead to performance degradation or necessitate substantial continued pre-training costs to restore performance. Based on the analysis of attention redundancy, we design a Decoupled-Head Attention (DHA) mechanism. DHA adaptively configures group sharing for key heads and value heads across various layers, achieving a better balance between performance and efficiency. Inspired by the observation of clustering similar heads, we propose to progressively transform the MHA checkpoint into the DHA model through linear fusion of similar head parameters step by step, retaining the parametric knowledge of the MHA checkpoint. We construct DHA models by transforming various scales of MHA checkpoints given target head budgets. Our experiments show that DHA remarkably requires a mere 0.25\% of the original model's pre-training budgets to achieve 97.6\% of performance while saving 75\% of KV cache. Compared to Group-Query Attention (GQA), DHA achieves a 5$\times$ training acceleration, a maximum of 13.93\% performance improvement under 0.01\% pre-training budget, and 4\% relative improvement under 0.05\% pre-training budget.

Beyond Gait: Learning Knee Angle for Seamless Prosthesis Control in Multiple Scenarios

Deep learning models have become a powerful tool in knee angle estimation for lower limb prostheses, owing to their adaptability across various gait phases and locomotion modes. Current methods utilize Multi-Layer Perceptrons (MLP), Long-Short Term Memory Networks (LSTM), and Convolutional Neural Networks (CNN), predominantly analyzing motion information from the thigh. Contrary to these approaches, our study introduces a holistic perspective by integrating whole-body movements as inputs. We propose a transformer-based probabilistic framework, termed the Angle Estimation Probabilistic Model (AEPM), that offers precise angle estimations across extensive scenarios beyond walking. AEPM achieves an overall RMSE of 6.70 degrees, with an RMSE of 3.45 degrees in walking scenarios. Compared to the state of the art, AEPM has improved the prediction accuracy for walking by 11.31%. Our method can achieve seamless adaptation between different locomotion modes. Also, this model can be utilized to analyze the synergy between the knee and other joints. We reveal that the whole body movement has valuable information for knee movement, which can provide insights into designing sensors for prostheses. The code is available at https://github.com/penway/Beyond-Gait-AEPM.

Integrated Sensing, Communication, and Powering (ISCAP): Towards Multi-functional 6G Wireless Networks

This article presents a novel multi-functional system for a sixth-generation (6G) wireless network with integrated sensing, communication, and powering (ISCAP), which unifies integrated sensing and communication (ISAC) and wireless information and power transfer (WIPT) techniques. The multi-functional ISCAP network promises to enhance resource utilization efficiency, reduce network costs, and improve overall performance through versatile operational modes. Specifically, a multi-functional base station (BS) can enable multi-functional transmission, by exploiting the same radio signals to perform target/environment sensing, wireless communication, and wireless power transfer (WPT), simultaneously. Besides, the three functions can be intelligently coordinated to pursue mutual benefits,i.e., wireless sensing can be leveraged to enable light-training or even training-free WIPT by providing side-channel information, and the BS can utilize WPT to wirelessly charge low-power devices for ensuring sustainable ISAC. Furthermore, multiple multi-functional BSs can cooperate in both transmission and reception phases for efficient interference management, multi-static sensing, and distributed energy beamforming. For these operational modes, we discuss the technical challenges and potential solutions, particularly focusing on the fundamental performance tradeoff limits, transmission protocol design, as well as waveform and beamforming optimization. Finally, interesting research directions are identified.