Search

Extending the functionality of the Transformer model to accommodate longer sequence lengths has become a critical challenge. This extension is crucial not only for improving tasks such as language translation and long-context processing but also for enabling novel applications like chatbots, code generation, and multimedia content creation. The primary obstacle is the self-attention mechanism, which scales quadratically with sequence length in terms of computation time and memory requirements. LongLoRA proposed shifted sparse attention (S\(^2\)-Attn), effectively enabling context extension and leading to non-trivial computation savings with similar performance to fine-tuning with vanilla attention. However, LongLoRA is still not as efficient as vanilla attention, reaching only 39\% of the perplexity improvement compared to full attention. This inefficiency is due to the cyclic shift applied within different attention head patterns, causing either chaos in the attention head structure or unnecessary information exchange between token groups. To address these issues, We propose \textbf{SinkLoRA}, which features better work partitioning. Specifically, (1) we developed SF-Attn with a segmentation and reassembly algorithm to proportionally return cyclically shifted groups of attention heads to their un-shifted state together with global attention of "sink attention tokens", achieving 92\% of the perplexity improvement compared to full attention after fine tuning, and (2) applied a SOTA KV cache compression algorithm H$_2$O to accelerate inference. Furthermore, We conducted supervised fine-tuning with SinkLoRA using a self collected LongAlpaca-plus dataset. All our code, models, datasets, and demos are available at \url{https://github.com/Dexter-GT-86/SinkLoRA}., Comment: A rethinking of Short Shifted Attention

Click-through rate (CTR) prediction tasks play a pivotal role in real-world applications, particularly in recommendation systems and online advertising. A significant research branch in this domain focuses on user behavior modeling. Current research predominantly centers on modeling co-occurrence relationships between the target item and items previously interacted with by users in their historical data. However, this focus neglects the intricate modeling of user behavior patterns. In reality, the abundance of user interaction records encompasses diverse behavior patterns, indicative of a spectrum of habitual paradigms. These patterns harbor substantial potential to significantly enhance CTR prediction performance. To harness the informational potential within user behavior patterns, we extend Target Attention (TA) to Target Pattern Attention (TPA) to model pattern-level dependencies. Furthermore, three critical challenges demand attention: the inclusion of unrelated items within behavior patterns, data sparsity in behavior patterns, and computational complexity arising from numerous patterns. To address these challenges, we introduce the Deep Pattern Network (DPN), designed to comprehensively leverage information from user behavior patterns. DPN efficiently retrieves target-related user behavior patterns using a target-aware attention mechanism. Additionally, it contributes to refining user behavior patterns through a pre-training paradigm based on self-supervised learning while promoting dependency learning within sparse patterns. Our comprehensive experiments, conducted across three public datasets, substantiate the superior performance and broad compatibility of DPN., Comment: 12 pages, 10 figures, accepted by SIGIR2024

The next generation wireless communication networks are required to support high-mobility scenarios, such as reliable data transmission for high-speed railways. Nevertheless, widely utilized multi-carrier modulation, the orthogonal frequency division multiplex (OFDM), cannot deal with the severe Doppler spread brought by high mobility. To address this problem, some new modulation schemes, e.g. orthogonal time frequency space and affine frequency division multiplexing, have been proposed with different design criteria from OFDM, which promote reliability with the cost of extremely high implementation complexity. On the other hand, end-to-end systems achieve excellent gains by exploiting neural networks to replace traditional transmitters and receivers, but have to retrain and update continually with channel varying. In this paper, we propose the Modem Network (ModNet) to design a novel modem scheme. Compared with end-to-end systems, channels are directly fed into the network and we can directly get a modem scheme through ModNet. Then, the Tri-Phase training strategy is proposed, which mainly utilizes the siamese structure to unify the learned modem scheme without retraining frequently faced up with time-varying channels. Simulation results show the proposed modem scheme outperforms OFDM systems under different highmobility channel statistics., Comment: 6 pages, 4 figures, accepted by ICC 2024 Workshop - APATN

Click-Through Rate (CTR) prediction, crucial in applications like recommender systems and online advertising, involves ranking items based on the likelihood of user clicks. User behavior sequence modeling has marked progress in CTR prediction, which extracts users' latent interests from their historical behavior sequences to facilitate accurate CTR prediction. Recent research explores using implicit feedback sequences, like unclicked records, to extract diverse user interests. However, these methods encounter key challenges: 1) temporal misalignment due to disparate sequence time ranges and 2) the lack of fine-grained interaction among feedback sequences. To address these challenges, we propose a novel framework called TEM4CTR, which ensures temporal alignment among sequences while leveraging auxiliary feedback information to enhance click behavior at the item level through a representation projection mechanism. Moreover, this projection-based information transfer module can effectively alleviate the negative impact of irrelevant or even potentially detrimental components of the auxiliary feedback information on the learning process of click behavior. Comprehensive experiments on public and industrial datasets confirm the superiority and effectiveness of TEM4CTR, showcasing the significance of temporal alignment in multi-feedback modeling., Comment: 11 pages, 5 figures

Multi-behavior recommendation algorithms aim to leverage the multiplex interactions between users and items to learn users' latent preferences. Recent multi-behavior recommendation frameworks contain two steps: fusion and prediction. In the fusion step, advanced neural networks are used to model the hierarchical correlations between user behaviors. In the prediction step, multiple signals are utilized to jointly optimize the model with a multi-task learning (MTL) paradigm. However, recent approaches have not addressed the issue caused by imbalanced data distribution in the fusion step, resulting in the learned relationships being dominated by high-frequency behaviors. In the prediction step, the existing methods use a gate mechanism to directly aggregate expert information generated by coupling input, leading to negative information transfer. To tackle these issues, we propose a Parallel Knowledge Enhancement Framework (PKEF) for multi-behavior recommendation. Specifically, we enhance the hierarchical information propagation in the fusion step using parallel knowledge (PKF). Meanwhile, in the prediction step, we decouple the representations to generate expert information and introduce a projection mechanism during aggregation to eliminate gradient conflicts and alleviate negative transfer (PME). We conduct comprehensive experiments on three real-world datasets to validate the effectiveness of our model. The results further demonstrate the rationality and effectiveness of the designed PKF and PME modules. The source code and datasets are available at https://github.com/MC-CV/PKEF., Comment: Accepted by CIKM 2023

In massive multiple-input multiple-output (MIMO) systems under the frequency division duplexing (FDD) mode, the user equipment (UE) needs to feed channel state information (CSI) back to the base station (BS). Though deep learning approaches have made a hit in the CSI feedback problem, whether they can remain excellent in actual environments needs to be further investigated. In this letter, we point out that the real-time dataset in application often has the domain gap from the training dataset caused by the time delay. To bridge the gap, we propose bubble-shift (B-S) data augmentation, which attempts to offset performance degradation by changing the delay and remaining the channel information as much as possible. Moreover, random-generation (R-G) data augmentation is especially proposed for outdoor scenarios due to the complex distribution of its channels. It generalizes the characteristics of the channel matrix and alleviates the over-fitting problem. Simulation results show that the proposed data augmentation boosts the robustness of networks in both indoor and outdoor environments. The open source codes are available at https://github.com/zhanghy23/CRNet-Aug.

Sequential recommendation (SR) plays an important role in personalized recommender systems because it captures dynamic and diverse preferences from users' real-time increasing behaviors. Unlike the standard autoregressive training strategy, future data (also available during training) has been used to facilitate model training as it provides richer signals about user's current interests and can be used to improve the recommendation quality. However, these methods suffer from a severe training-inference gap, i.e., both past and future contexts are modeled by the same encoder when training, while only historical behaviors are available during inference. This discrepancy leads to potential performance degradation. To alleviate the training-inference gap, we propose a new framework DualRec, which achieves past-future disentanglement and past-future mutual enhancement by a novel dual network. Specifically, a dual network structure is exploited to model the past and future context separately. And a bi-directional knowledge transferring mechanism enhances the knowledge learnt by the dual network. Extensive experiments on four real-world datasets demonstrate the superiority of our approach over baseline methods. Besides, we demonstrate the compatibility of DualRec by instantiating using RNN, Transformer, and filter-MLP as backbones. Further empirical analysis verifies the high utility of modeling future contexts under our DualRec framework. The code of DualRec is publicly available at https://github.com/zhy99426/DualRec., Comment: Accept by CIKM 2022

Benefiting from the event-driven and sparse spiking characteristics of the brain, spiking neural networks (SNNs) are becoming an energy-efficient alternative to artificial neural networks (ANNs). However, the performance gap between SNNs and ANNs has been a great hindrance to deploying SNNs ubiquitously for a long time. To leverage the full potential of SNNs, we study the effect of attention mechanisms in SNNs. We first present our idea of attention with a plug-and-play kit, termed the Multi-dimensional Attention (MA). Then, a new attention SNN architecture with end-to-end training called "MA-SNN" is proposed, which infers attention weights along the temporal, channel, as well as spatial dimensions separately or simultaneously. Based on the existing neuroscience theories, we exploit the attention weights to optimize membrane potentials, which in turn regulate the spiking response in a data-dependent way. At the cost of negligible additional parameters, MA facilitates vanilla SNNs to achieve sparser spiking activity, better performance, and energy efficiency concurrently. Experiments are conducted in event-based DVS128 Gesture/Gait action recognition and ImageNet-1k image classification. On Gesture/Gait, the spike counts are reduced by 84.9%/81.6%, and the task accuracy and energy efficiency are improved by 5.9%/4.7% and 3.4$\times$/3.2$\times$. On ImageNet-1K, we achieve top-1 accuracy of 75.92% and 77.08% on single/4-step Res-SNN-104, which are state-of-the-art results in SNNs. To our best knowledge, this is for the first time, that the SNN community achieves comparable or even better performance compared with its ANN counterpart in the large-scale dataset. Our work lights up SNN's potential as a general backbone to support various applications for SNNs, with a great balance between effectiveness and efficiency., Comment: 18 pages, 8 figures, Under Review

We present a finite-temperature density functional theory investigation of the nonequilibrium transient electronic structure of warm dense Li, Al, Cu, and Au created by laser excitation. Photons excite electrons either from the inner shell orbitals or from the valence bands according to the photon energy, and give rise to isochoric heating of the sample. Localized states related to the 3d orbital are observed for Cu when the hole lies in the inner shell 3s orbital. The electrical conductivity for these materials at nonequilibrium states is calculated using the Kubo-Greenwood formula. The change of the electrical conductivity, compared to the equilibrium state, is different for the case of holes in inner shell orbitals or the valence band. This is attributed to the competition of two factors: the shift of the orbital energies due to reduced screening of core electrons, and the increase of chemical potential due to the excitation of electrons. The finite temperature effect of both the electrons and the ions on the electrical conductivity is discussed in detail. This work is helpful to better understand the physics of laser excitation experiments of warm dense matter., Comment: 111 pages, 9 figures