EvoP: Robust LLM Inference via Evolutionary Pruning

Large Language Models (LLMs) have achieved remarkable success in natural language processing tasks, but their massive size and computational demands hinder their deployment in resource-constrained environments. Existing structured pruning methods address this issue by removing redundant structures (e.g., elements, channels, layers) from the model. However, these methods employ a heuristic pruning strategy, which leads to suboptimal performance. Besides, they also ignore the data characteristics when pruning the model. To overcome these limitations, we propose EvoP, an evolutionary pruning framework for robust LLM inference. EvoP first presents a cluster-based calibration dataset sampling (CCDS) strategy for creating a more diverse calibration dataset. EvoP then introduces an evolutionary pruning pattern searching (EPPS) method to find the optimal pruning pattern. Compared to existing structured pruning techniques, EvoP achieves the best performance while maintaining the best efficiency. Experiments across different LLMs and different downstream tasks validate the effectiveness of the proposed EvoP, making it a practical and scalable solution for deploying LLMs in real-world applications.

GeneQuery: A General QA-based Framework for Spatial Gene Expression Predictions from Histology Images

Gene expression profiling provides profound insights into molecular mechanisms, but its time-consuming and costly nature often presents significant challenges. In contrast, whole-slide hematoxylin and eosin (H&E) stained histological images are readily accessible and allow for detailed examinations of tissue structure and composition at the microscopic level. Recent advancements have utilized these histological images to predict spatially resolved gene expression profiles. However, state-of-the-art works treat gene expression prediction as a multi-output regression problem, where each gene is learned independently with its own weights, failing to capture the shared dependencies and co-expression patterns between genes. Besides, existing works can only predict gene expression values for genes seen during training, limiting their ability to generalize to new, unseen genes. To address the above limitations, this paper presents GeneQuery, which aims to solve this gene expression prediction task in a question-answering (QA) manner for better generality and flexibility. Specifically, GeneQuery takes gene-related texts as queries and whole-slide images as contexts and then predicts the queried gene expression values. With such a transformation, GeneQuery can implicitly estimate the gene distribution by introducing the gene random variable. Besides, the proposed GeneQuery consists of two architecture implementations, i.e., spot-aware GeneQuery for capturing patterns between images and gene-aware GeneQuery for capturing patterns between genes. Comprehensive experiments on spatial transcriptomics datasets show that the proposed GeneQuery outperforms existing state-of-the-art methods on known and unseen genes. More results also demonstrate that GeneQuery can potentially analyze the tissue structure.

FAST: A Dual-tier Few-Shot Learning Paradigm for Whole Slide Image Classification

The expensive fine-grained annotation and data scarcity have become the primary obstacles for the widespread adoption of deep learning-based Whole Slide Images (WSI) classification algorithms in clinical practice. Unlike few-shot learning methods in natural images that can leverage the labels of each image, existing few-shot WSI classification methods only utilize a small number of fine-grained labels or weakly supervised slide labels for training in order to avoid expensive fine-grained annotation. They lack sufficient mining of available WSIs, severely limiting WSI classification performance. To address the above issues, we propose a novel and efficient dual-tier few-shot learning paradigm for WSI classification, named FAST. FAST consists of a dual-level annotation strategy and a dual-branch classification framework. Firstly, to avoid expensive fine-grained annotation, we collect a very small number of WSIs at the slide level, and annotate an extremely small number of patches. Then, to fully mining the available WSIs, we use all the patches and available patch labels to build a cache branch, which utilizes the labeled patches to learn the labels of unlabeled patches and through knowledge retrieval for patch classification. In addition to the cache branch, we also construct a prior branch that includes learnable prompt vectors, using the text encoder of visual-language models for patch classification. Finally, we integrate the results from both branches to achieve WSI classification. Extensive experiments on binary and multi-class datasets demonstrate that our proposed method significantly surpasses existing few-shot classification methods and approaches the accuracy of fully supervised methods with only 0.22$\%$ annotation costs. All codes and models will be publicly available on https://github.com/fukexue/FAST.

Retrieval-Augmented Generation for Natural Language Processing: A Survey

Large language models (LLMs) have demonstrated great success in various fields, benefiting from their huge amount of parameters that store knowledge. However, LLMs still suffer from several key issues, such as hallucination problems, knowledge update issues, and lacking domain-specific expertise. The appearance of retrieval-augmented generation (RAG), which leverages an external knowledge database to augment LLMs, makes up those drawbacks of LLMs. This paper reviews all significant techniques of RAG, especially in the retriever and the retrieval fusions. Besides, tutorial codes are provided for implementing the representative techniques in RAG. This paper further discusses the RAG training, including RAG with/without datastore update. Then, we introduce the application of RAG in representative natural language processing tasks and industrial scenarios. Finally, this paper discusses the future directions and challenges of RAG for promoting its development.

RAEE: A Training-Free Retrieval-Augmented Early Exiting Framework for Efficient Inference

Deploying large language model inference remains challenging due to their high computational overhead. Early exiting accelerates model inference by adaptively reducing the number of inference layers. Existing methods require training internal classifiers to determine whether to exit at each intermediate layer. However, such classifier-based early exiting frameworks require significant effort to design and train the classifiers. To address these limitations, this paper proposes RAEE, a training-free Retrieval-Augmented Early Exiting framework for efficient inference. First, this paper demonstrates that the early exiting problem can be modeled as a distribution prediction problem, where the distribution is approximated using similar data's existing information. Next, the paper details the process of collecting existing information to build the retrieval database. Finally, based on the pre-built retrieval database, RAEE leverages the retrieved similar data's exiting information to guide the backbone model to exit at the layer, which is predicted by the approximated distribution. Experimental results demonstrate that the proposed RAEE can significantly accelerate inference. RAEE also achieves state-of-the-art zero-shot performance on 8 classification tasks.

Improving Natural Language Understanding with Computation-Efficient Retrieval Representation Fusion

Retrieval-based augmentations that aim to incorporate knowledge from an external database into language models have achieved great success in various knowledge-intensive (KI) tasks, such as question-answering and text generation. However, integrating retrievals in non-knowledge-intensive (NKI) tasks, such as text classification, is still challenging. Existing works focus on concatenating retrievals to inputs as context to form the prompt-based inputs. Unfortunately, such methods require language models to have the capability to handle long texts. Besides, inferring such concatenated data would also consume a significant amount of computational resources. To solve these challenges, we propose \textbf{ReFusion} in this paper, a computation-efficient \textbf{Re}trieval representation \textbf{Fusion} with neural architecture search. The main idea is to directly fuse the retrieval representations into the language models. Specifically, we first propose an online retrieval module that retrieves representations of similar sentences. Then, we present a retrieval fusion module including two effective ranking schemes, i.e., reranker-based scheme and ordered-mask-based scheme, to fuse the retrieval representations with hidden states. Furthermore, we use Neural Architecture Search (NAS) to seek the optimal fusion structure across different layers. Finally, we conduct comprehensive experiments, and the results demonstrate our ReFusion can achieve superior and robust performance on various NKI tasks.

Distributed UAV Swarm Augmented Wideband Spectrum Sensing Using Nyquist Folding Receiver

Distributed unmanned aerial vehicle (UAV) swarms are formed by multiple UAVs with increased portability, higher levels of sensing capabilities, and more powerful autonomy. These features make them attractive for many recent applica-tions, potentially increasing the shortage of spectrum resources. In this paper, wideband spectrum sensing augmented technology is discussed for distributed UAV swarms to improve the utilization of spectrum. However, the sub-Nyquist sampling applied in existing schemes has high hardware complexity, power consumption, and low recovery efficiency for non-strictly sparse conditions. Thus, the Nyquist folding receiver (NYFR) is considered for the distributed UAV swarms, which can theoretically achieve full-band spectrum detection and reception using a single analog-to-digital converter (ADC) at low speed for all circuit components. There is a focus on the sensing model of two multichannel scenarios for the distributed UAV swarms, one with a complete functional receiver for the UAV swarm with RIS, and another with a decentralized UAV swarm equipped with a complete functional receiver for each UAV element. The key issue is to consider whether the application of RIS technology will bring advantages to spectrum sensing and the data fusion problem of decentralized UAV swarms based on the NYFR architecture. Therefore, the property for multiple pulse reconstruction is analyzed through the Gershgorin circle theorem, especially for very short pulses. Further, the block sparse recovery property is analyzed for wide bandwidth signals. The proposed technology can improve the processing capability for multiple signals and wide bandwidth signals while reducing interference from folded noise and subsampled harmonics. Experiment results show augmented spectrum sensing efficiency under non-strictly sparse conditions.

Wideband Power Spectrum Sensing: a Fast Practical Solution for Nyquist Folding Receiver

The limited availability of spectrum resources has been growing into a critical problem in wireless communications, remote sensing, and electronic surveillance, etc. To address the high-speed sampling bottleneck of wideband spectrum sensing, a fast and practical solution of power spectrum estimation for Nyquist folding receiver (NYFR) is proposed in this paper. The NYFR architectures is can theoretically achieve the full-band signal sensing with a hundred percent of probability of intercept. But the existing algorithm is difficult to realize in real-time due to its high complexity and complicated calculations. By exploring the sub-sampling principle inherent in NYFR, a computationally efficient method is introduced with compressive covariance sensing. That can be efficient implemented via only the non-uniform fast Fourier transform, fast Fourier transform, and some simple multiplication operations. Meanwhile, the state-of-the-art power spectrum reconstruction model for NYFR of time-domain and frequency-domain is constructed in this paper as a comparison. Furthermore, the computational complexity of the proposed method scales linearly with the Nyquist-rate sampled number of samples and the sparsity of spectrum occupancy. Simulation results and discussion demonstrate that the low complexity in sampling and computation is a more practical solution to meet the real-time wideband spectrum sensing applications.

LightSeq2: Accelerated Training for Transformer-based Models on GPUs

Transformer-based models have proven to be powerful in many natural language, computer vision, and speech recognition applications. It is expensive to train these types of models due to unfixed input length, complex computation, and large numbers of parameters. Existing systems either only focus on efficient inference or optimize only BERT-like encoder models. In this paper, we present LightSeq2, a system for efficient training of Transformer-based models on GPUs. We propose a series of GPU optimization techniques tailored to computation flow and memory access patterns of neural layers in Transformers. LightSeq2 supports a variety of network architectures, including BERT (encoder-only), GPT (decoder-only), and Transformer (encoder-decoder). Our experiments on GPUs with varying models and datasets show that LightSeq2 is 1.4-3.5x faster than previous systems. In particular, it gains 308% training speedup compared with existing systems on a large public machine translation benchmark (WMT14 English-German).

LightSeq: A High Performance Inference Library for Sequence Processing and Generation

LightSeq is a high performance inference library for sequence processing and generation implemented in CUDA. To our best knowledge, this is the first open-source inference library which fully supports highly efficient computation of modern NLP models such as BERT, GPT, Transformer, etc. This library is efficient, functional and convenient. A demo usage can be found here: https://github.com/bytedance/lightseq/blob/master/example.