论文检索

Deep Learning · Robustness

Amirabbas Afzali, Borna khodabandeh, Ali Rasekh, Mahyar JafariNodeh, Sepehr Ranjbar, Simon Gottschalk

Contrastive learning models have demonstrated impressive abilities to capture semantic similarities by aligning representations in the embedding space. However, their performance can be limited by the quality of the training data and its inherent biases. While Preference Optimization (PO) methods such as Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimization (DPO) have been applied to align generative models with human preferences, their use in contrastive learning has yet to be explored. This paper introduces a novel method for training contrastive learning models using different PO methods to break down complex concepts. Our method systematically aligns model behavior with desired preferences, enhancing performance on the targeted task. In particular, we focus on enhancing model robustness against typographic attacks and inductive biases, commonly seen in contrastive vision-language models like CLIP. Our experiments demonstrate that models trained using PO outperform standard contrastive learning techniques while retaining their ability to handle adversarial challenges and maintain accuracy on other downstream tasks. This makes our method well-suited for tasks requiring fairness, robustness, and alignment with specific preferences. We evaluate our method for tackling typographic attacks on images and explore its ability to disentangle gender concepts and mitigate gender bias, showcasing the versatility of our approach.

Applications · Everything Else

Weihan Xu, Paul Pu Liang, Haven Kim, Julian McAuley, Taylor Berg-Kirkpatrick, Hao-Wen (Herman) Dong

Teasers are an effective tool for promoting content in entertainment, commercial and educational fields. However, creating an effective teaser for long videos is challenging for it requires long-range multimodal modeling capability for the input videos, while necessitating maintaining audiovisual alignments, managing scene transitions and preserving factual accuracy for the output teasers. Due to the lack of a publicly-available dataset, progress along this research direction has been hindered. In this work, we present DocumentaryNet, a collection of 1,269 documentaries paired with their teasers, featuring multimodal data streams of video, speech, music, sound effects and narrations. With DocumentaryNet, we propose a new two-stage system for generating teasers from long documentaries. The proposed TeaserGen system first generates the teaser narration from the transcribed narration from the documentary using a pretrained large language model, and then selects the most relevant visual content to accompany the generated narration through language-vision models. For narration-video matching, we explore two approaches: a pretraining-based model using pretrained contrastive language-vision models and a deep sequential model that learns the mapping between the narrations and visuals. Our experimental results show that the pretraining-based approach is more effective at identifying relevant visual content than directly trained deep autoregressive models.

Deep Learning · Large Language Models

Yuxian Gu, Li Dong, Hongning Wang, Yaru Hao, Qingxiu Dong, Furu Wei, Minlie Huang

This work investigates the selection of high-quality pre-training data from massive corpora to enhance LMs' capabilities for downstream usage. We formulate data selection as a generalized Optimal Control problem, which can be solved theoretically by Pontryagin's Maximum Principle (PMP), yielding a set of necessary conditions that characterize the relationship between optimal data selection and LM training dynamics.Based on these theoretical results, we introduce **P**MP-based **D**ata **S**election (**PDS**), a framework that approximates optimal data selection by solving the PMP conditions. In our experiments, we adopt PDS to select data from CommmonCrawl and show that the PDS-selected corpus accelerates the learning of LMs and constantly boosts their performance on a wide range of downstream tasks across various model sizes.Moreover, the benefits of PDS extend to ~400B models trained on ~10T tokens, as evidenced by the extrapolation of the test loss curves according to the Scaling Laws.PDS also improves data utilization when the pre-training data is limited, by reducing the data demand by 1.8 times, which helps mitigate the quick exhaustion of available web-crawled corpora. Our code, model, and data can be found at https://github.com/microsoft/LMOps/tree/main/data_selection.

Deep Learning · Large Language Models

Toshiaki Koike-Akino, Francesco Tonin, Yongtao Wu, Frank Zhengqing Wu, Leyla Naz Candogan, Volkan Cevher

This paper introduces Quantum-PEFT that leverages quantum computations for parameter-efficient fine-tuning (PEFT). Unlike other additive PEFT methods, such as low-rank adaptation (LoRA), Quantum-PEFT exploits an underlying full-rank yet surprisingly parameter efficient _quantum unitary parameterization_. With the use of Pauli parameterization, the number of trainable parameters grows only logarithmically with the ambient dimension, as opposed to linearly as in LoRA-based PEFT methods. Quantum-PEFT achieves vanishingly smaller number of trainable parameters than the lowest-rank LoRA as dimensions grow, enhancing parameter efficiency while maintaining a competitive performance. We apply Quantum-PEFT to several transfer learning benchmarks in language and vision, demonstrating significant advantages in parameter efficiency.

Optimization · Non-Convex

Frank Zhengqing Wu, Berfin Simsek, François Ged

In this paper, we study the loss landscape of one-hidden-layer neural networks with ReLU-like activation functions trained with the empirical squared loss using gradient descent (GD). We identify the stationary points of such networks, which significantly slow down loss decrease during training. To capture such points while accounting for the non-differentiability of the loss, the stationary points that we study are directional stationary points, rather than other notions like Clarke stationary points. We show that, if a stationary point does not contain "escape neurons", which are defined with first-order conditions, it must be a local minimum. Moreover, for the scalar-output case, the presence of an escape neuron guarantees that the stationary point is not a local minimum. Our results refine the description of the *saddle-to-saddle* training process starting from infinitesimally small (vanishing) initialization for shallow ReLU-like networks: By precluding the saddle escape types that previous works did not rule out, we advance one step closer to a complete picture of the entire dynamics. Moreover, we are also able to fully discuss how network embedding, which is to instantiate a narrower network with a wider network, reshapes the stationary points.

Applications · Computer Vision

Chaodong Xiao, Minghan Li, zhengqiang ZHANG, Deyu Meng, Lei Zhang

Selective state space models (SSMs), such as Mamba, highly excel at capturing long-range dependencies in 1D sequential data, while their applications to 2D vision tasks still face challenges. Current visual SSMs often convert images into 1D sequences and employ various scanning patterns to incorporate local spatial dependencies. However, these methods are limited in effectively capturing the complex image spatial structures and the increased computational cost caused by the lengthened scanning paths. To address these limitations, we propose Spatial-Mamba, a novel approach that establishes neighborhood connectivity directly in the state space. Instead of relying solely on sequential state transitions, we introduce a structure-aware state fusion equation, which leverages dilated convolutions to capture image spatial structural dependencies, significantly enhancing the flow of visual contextual information. Spatial-Mamba proceeds in three stages: initial state computation in a unidirectional scan, spatial context acquisition through structure-aware state fusion, and final state computation using the observation equation. Our theoretical analysis shows that Spatial-Mamba unifies the original Mamba and linear attention under the same matrix multiplication framework, providing a deeper understanding of our method. Experimental results demonstrate that Spatial-Mamba, even with a single scan, attains or surpasses the state-of-the-art SSM-based models in image classification, detection and segmentation. Source codes and trained models can be found at \url{ https://github.com/EdwardChasel/Spatial-Mamba }.

Applications · Neuroscience, Cognitive Science

Jan-Matthis Lueckmann, Alexander Immer, Alex Chen, Peter Li, Mariela Petkova, Nirmala Iyer, Luuk Hesselink, Aparna Dev, Gudrun Ihrke, Woohyun Park 等

Data-driven benchmarks have led to significant progress in key scientific modeling domains including weather and structural biology. Here, we introduce the Zebrafish Activity Prediction Benchmark (ZAPBench) to measure progress on the problem of predicting cellular-resolution neural activity throughout an entire vertebrate brain. The benchmark is based on a novel dataset containing 4d light-sheet microscopy recordings of over 70,000 neurons in a larval zebrafish brain, along with motion stabilized and voxel-level cell segmentations of these data that facilitate development of a variety of forecasting methods. Initial results from a selection of time series and volumetric video modeling approaches achieve better performance than naive baseline methods, but also show room for further improvement. The specific brain used in the activity recording is also undergoing synaptic-level anatomical mapping, which will enable future integration of detailed structural information into forecasting methods.

Theory · Learning Theory

Seongmin Lee, Marcel Boehme

The *missing mass* refers to the proportion of data points in an *unknown* population of classifier inputs that belong to classes *not* present in the classifier's training data, which is assumed to be a random sample from that unknown population.We find that *in expectation* the missing mass is entirely determined by the number $f_k$ of classes that *do* appear in the training data the same number of times *and an exponentially decaying error*.While this is the first precise characterization of the expected missing mass in terms of the sample, the induced estimator suffers from an impractically high variance. However, our theory suggests a large search space of nearly unbiased estimators that can be searched effectively and efficiently. Hence, we cast distribution-free estimation as an optimization problem to find a distribution-specific estimator with a minimized mean-squared error (MSE), given only the sample.In our experiments, our search algorithm discovers estimators that have a substantially smaller MSE than the state-of-the-art Good-Turing estimator. This holds for over 93\% of runs when there are at least as many samples as classes. Our estimators' MSE is roughly 80\% of the Good-Turing estimator's.

Deep Learning · Generative Models and Autoencoders

zhengqiang ZHANG, Ruihuang Li, Lei Zhang

While image generation with diffusion models has achieved a great success, generating images of higher resolution than the training size remains a challenging task due to the high computational cost. Current methods typically perform the entire sampling process at full resolution and process all frequency components simultaneously, contradicting with the inherent coarse-to-fine nature of latent diffusion models and wasting computations on processing premature high-frequency details at early diffusion stages. To address this issue, we introduce an efficient $\textbf{Fre}$quency-aware $\textbf{Ca}$scaded $\textbf{S}$ampling framework, $\textbf{FreCaS}$ in short, for higher-resolution image generation. FreCaS decomposes the sampling process into cascaded stages with gradually increased resolutions, progressively expanding frequency bands and refining the corresponding details. We propose an innovative frequency-aware classifier-free guidance (FA-CFG) strategy to assign different guidance strengths for different frequency components, directing the diffusion model to add new details in the expanded frequency domain of each stage. Additionally, we fuse the cross-attention maps of previous and current stages to avoid synthesizing unfaithful layouts. Experiments demonstrate that FreCaS significantly outperforms state-of-the-art methods in image quality and generation speed. In particular, FreCaS is about 2.86$\times$ and 6.07$\times$ faster than ScaleCrafter and DemoFusion in generating a 2048$\times$2048 image using a pretrained SDXL model and achieves an $\text{FID}_b$ improvement of 11.6 and 3.7, respectively. FreCaS can be easily extended to more complex models such as SD3. The source code of FreCaS can be found at https://github.com/xtudbxk/FreCaS.

Deep Learning · Large Language Models

Yifei Xing, Xiangyuan Lan, Ruiping Wang, Dongmei Jiang, Wenjun Huang, Zheng Qingfang, Yaowei Wang

Mamba-based architectures have shown to be a promising new direction for deep learning models owing to their competitive performance and sub-quadratic deployment speed. However, current Mamba multi-modal large language models (MLLM) are insufficient in extracting visual features, leading to imbalanced cross-modal alignment between visual and textural latents, negatively impacting performance on multi-modal tasks. In this work, we propose Empowering Multi-modal Mamba with Structural and Hierarchical Alignment (EMMA), which enables the MLLM to extract fine-grained visual information. Specifically, we propose a pixel-wise alignment module to autoregressively optimize the learning and processing of spatial image-level features along with textual tokens, enabling structural alignment at the image level. In addition, to prevent the degradation of visual information during the cross-model alignment process, we propose a multi-scale feature fusion (MFF) module to combine multi-scale visual features from intermediate layers, enabling hierarchical alignment at the feature level. Extensive experiments are conducted across a variety of multi-modal benchmarks. Our model shows lower latency than other Mamba-based MLLMs and is nearly four times faster than transformer-based MLLMs of similar scale during inference. Due to better cross-modal alignment, our model exhibits lower degrees of hallucination and enhanced sensitivity to visual details, which manifests in superior performance across diverse multi-modal benchmarks. Code provided at https://github.com/xingyifei2016/EMMA.

Deep Learning · Large Language Models

Rana Shahout, Eran Malach, Chunwei Liu, Weifan Jiang, Minlan Yu, Michael Mitzenmacher

Efficient scheduling is crucial for interactive Large Language Model (LLM) applications, where low request completion time directly impacts user engagement. Size-based scheduling algorithms like Shortest Remaining Process Time (SRPT) aim to reduce average request completion time by leveraging known or estimated request sizes and allowing preemption by incoming jobs with shorter service times. However, two main challenges arise when applying size-based scheduling to LLM systems. First, accurately predicting output lengths from prompts is challenging and often resource-intensive, making it impractical for many systems. As a result, the state-of-the-art LLM systems default to first-come, first-served scheduling, which can lead to head-of-line blocking and reduced system efficiency. Second, preemption introduces extra memory overhead to LLM systems as they must maintain intermediate states for unfinished (preempted) requests.In this paper, we propose TRAIL, a method to obtain output predictions from the target LLM itself. After generating each output token, we recycle the embedding of its internal structure as input for a lightweight classifier that predicts the remaining length for each running request. Using these predictions, we propose a prediction-based SRPT variant with limited preemption designed to account for memory overhead in LLM systems. This variant allows preemption early in request execution when memory consumption is low but restricts preemption as requests approach completion to optimize resource utilization. On the theoretical side, we derive a closed-form formula for this SRPT variant in an M/G/1 queue model, which demonstrates its potential value. In our system, we implement this preemption policy alongside our embedding-based prediction method. Our refined predictions from layer embeddings achieve 2.66x lower mean absolute error compared to BERT predictions from sequence prompts. TRAIL achieves 1.66x to 2.01x lower mean latency on the Alpaca dataset and 1.76x to 24.07x lower mean time to the first token compared to the state-of-the-art serving system.

Xingyao Wang, Boxuan Li, Yufan Song, Frank F Xu, Xiangru Tang, Mingchen Zhuge, Jiayi Pan, Yueqi Song, Bowen Li, Jaskirat Singh 等

Software is one of the most powerful tools that we humans have at our disposal; it allows a skilled programmer to interact with the world in complex and profound ways. At the same time, thanks to improvements in large language models (LLMs), there has also been a rapid development in AI agents that interact with and effect change in their surrounding environments. In this paper, we introduce OpenHands, a platform for the development of powerful and flexible AI agents that interact with the world in similar ways to a human developer: by writing code, interacting with a command line, and browsing the web. We describe how the platform allows for the implementation of new agents, utilization of various LLMs, safe interaction with sandboxed environments for code execution, and incorporation of evaluation benchmarks. Based on our currently incorporated benchmarks, we perform an evaluation of agents over 13 challenging tasks, including software engineering (e.g., SWE-Bench) and web browsing (e.g., WebArena), amongst others. Released under the permissive MIT license, OpenHands is a community project spanning academia and industry with more than 2K contributions from over 186 contributors in less than six months of development, and will improve going forward.

Social Aspects · Privacy-preserving Statistics and Machine Learning

Yancheng Zhang, Jiaqi Xue, Mengxin Zheng, Mimi Xie, Mingzhe Zhang, Lei Jiang, Qian Lou

Private Transformer inference using cryptographic protocols offers promising solutions for privacy-preserving machine learning; however, it still faces significant runtime overhead (efficiency issues) and challenges in handling long-token inputs (scalability issues). We observe that the Transformer's operational complexity scales quadratically with the number of input tokens, making it essential to reduce the input token length. Notably, each token varies in importance, and many inputs contain redundant tokens. Additionally, prior private inference methods that rely on high-degree polynomial approximations for non-linear activations are computationally expensive. Therefore, reducing the polynomial degree for less important tokens can significantly accelerate private inference. Building on these observations, we propose \textit{CipherPrune}, an efficient and scalable private inference framework that includes a secure encrypted token pruning protocol, a polynomial reduction protocol, and corresponding Transformer network optimizations. At the protocol level, encrypted token pruning adaptively removes unimportant tokens from encrypted inputs in a progressive, layer-wise manner. Additionally, encrypted polynomial reduction assigns lower-degree polynomials to less important tokens after pruning, enhancing efficiency without decryption. At the network level, we introduce protocol-aware network optimization via a gradient-based search to maximize pruning thresholds and polynomial reduction conditions while maintaining the desired accuracy. Our experiments demonstrate that CipherPrune reduces the execution overhead of private Transformer inference by approximately $6.1\times$ for 128-token inputs and $10.6\times$ for 512-token inputs, compared to previous methods, with only a marginal drop in accuracy. The code is publicly available at https://github.com/UCF-Lou-Lab-PET/cipher-prune-inference.

Applications · Computer Vision

Jilan Xu, Yifei Huang, Baoqi Pei, Junlin Hou, Qingqiu Li, Guo Chen, Yuejie Zhang, Rui Feng, Weidi Xie

Generating videos in the first-person perspective has broad application prospects in the field of augmented reality and embodied intelligence.In this work, we explore the cross-view video prediction task, where given an exo-centric video, the first frame of the corresponding ego-centric video, and textual instructions, the goal is to generate future frames of the ego-centric video. Inspired by the notion that hand-object interactions (HOI) in ego-centric videos represent the primary intentions and actions of the current actor, we present EgoExo-Gen that explicitly models the hand-object dynamics for cross-view video prediction. EgoExo-Gen consists of two stages. First, we design a cross-view HOI mask prediction model that anticipates the HOI masks in future ego-frames by modeling the spatio-temporal ego-exo correspondence. Next, we employ a video diffusion model to predict future ego-frames using the first ego-frame and textual instructions, while incorporating the HOI masks as structural guidance to enhance prediction quality.To facilitate training, we develop a fully automated pipeline to generate pseudo HOI masks for both ego- and exo-videos by exploiting vision foundation models. Extensive experiments demonstrate that our proposed EgoExo-Gen achieves better prediction performance compared to previous video prediction models on the public Ego-Exo4D and H2O benchmark datasets, with the HOI masks significantly improving the generation of hands and interactive objects in the ego-centric videos.

Deep Learning · Attention Mechanisms

Guoxia Wang, Jinle Zeng, Xiyuan Xiao, Siming Wu, Jiabin Yang, Lujing Zheng, Zeyu Chen, Jiang Bian, Dianhai Yu, Haifeng Wang

The computational and memory demands of vanilla attention scale quadratically with the sequence length $N$, posing significant challenges for processing long sequences in Transformer models. FlashAttention alleviates these challenges by eliminating the $\mathcal{O}(N^2)$ memory dependency and reducing attention latency through IO-aware memory optimizations. However, its native support for certain attention mask types is limited, and it does not inherently accommodate more complex masking requirements. Previous approaches resort to using dense masks with $\mathcal{O}(N^2)$ memory complexity, leading to inefficiencies. In this paper, we propose \ours{}, an extension of FlashAttention that introduces a column-wise sparse representation of attention masks. This approach efficiently represents a wide range of mask types and facilitates the development of optimized kernel implementations. By adopting this novel representation, \ours{} achieves linear memory complexity $\mathcal{O}(N)$, making it suitable for modeling long-context sequences. Moreover, this representation enables kernel optimizations that eliminate unnecessary computations by leveraging sparsity in the attention mask, without sacrificing computational accuracy, resulting in higher computational efficiency. We evaluate \ours{}'s performance in fine-tuning and alignment training of LLMs such as SFT, LoRA, DPO, and RM. \ours{} achieves significant throughput improvements, with end-to-end speedups ranging from 1.65x to 3.22x compared to existing FlashAttention dense method. Additionally, our kernel-level comparisons demonstrate that \ours{} surpasses the latest counterpart, FlexAttention, by 12.1\% to 60.7\% in terms of kernel TFLOPs/s, achieving 37.8\% to 62.3\% of the theoretical maximum FLOPs/s on the A100 GPU. The code is open-sourced on PaddlePaddle\footnote{\url{https://github.com/PaddlePaddle/Paddle}} and integrated into PaddleNLP\footnote{\url{https://github.com/PaddlePaddle/PaddleNLP}}, supporting models with over 100 billion parameters for contexts extending up to 128K tokens.

Miscellaneous Aspects of Machine Learning · Unsupervised and Semi-supervised Learning

Wei Dai, Jicong Fan

Unsupervised anomaly detection (UAD) has important applications in diverse fields such as manufacturing industry and medical diagnosis. In the past decades, although numerous insightful and effective UAD methods have been proposed, it remains a huge challenge to tune the hyper-parameters of each method and select the most appropriate method among many candidates for a specific dataset, due to the absence of labeled anomalies in the training phase of UAD methods and the high diversity of real datasets. In this work, we aim to address this challenge, so as to make UAD more practical and reliable. We propose two internal evaluation metrics, relative-top-median and expected-anomaly-gap, and one semi-internal evaluation metric, normalized pseudo discrepancy (NPD), as surrogate functions of the expected model performance on unseen test data. For instance, NPD measures the discrepancy between the anomaly scores of a validation set drawn from the training data and a validation set drawn from an isotropic Gaussian. NPD is simple and hyper-parameter-free and is able to compare different UAD methods, and its effectiveness is theoretically analyzed. We integrate the three metrics with Bayesian optimization to effectively optimize the hyper-parameters of UAD models. Extensive experiments on 38 datasets show the effectiveness of our methods.

Deep Learning · Large Language Models

Kenneth Enevoldsen, Isaac Chung, Imene Kerboua, Márton Kardos, Ashwin Mathur, David Stap, Jay Gala, Wissam Siblini, Dominik Krzemiński, Genta Winata 等

Text embeddings are typically evaluated on a narrow set of tasks, limited in terms of languages, domains, and task types. To circumvent this limitation and to provide a more comprehensive evaluation, we introduce the Massive Multilingual Text Embedding Benchmark (MMTEB) -- a large-scale community-driven initiative expanding MTEB to over 500 quality-controlled evaluation tasks across 1,000+ languages. MMTEB includes a wide range of challenging novel tasks such as instruction following, long-document retrieval, and code retrieval, and represents the largest multilingual collection of evaluation tasks for embedding models to date. We use this collection to construct multiple highly multilingual benchmarks. We evaluate a representative set of models on these benchmarks.Our findings indicate that, while LLM-based models can achieve state-of-the-art performance on a subset of languages, the best-performing publicly available model across languages is the notably smaller, multilingual-e5-large-instruct.Massive benchmarks often impose high computational demands, limiting accessibility, particularly for low-resource communities. To address this, we downsample tasks based on inter-task correlation (i.e., selecting only a diverse set of tasks) while preserving relative rankings.We further optimize tasks such as retrieval by sampling hard negatives, creating smaller but effective splits. These optimizations allow us to introduce benchmarks at a significantly lower computational cost. For instance, we introduce a new zero-shot English benchmark that maintains a similar ordering at a fraction of the cost.

Deep Learning · Large Language Models

Jaehyung Seo, Heuiseok Lim

Recent researchers and companies have been developing large language models (LLMs) specifically designed for particular purposes and have achieved significant advancements in various natural language processing tasks. However, LLMs are still prone to generating hallucinations—results that are unfaithful or inconsistent with the given input. As a result, the need for datasets to evaluate and demonstrate the hallucination detection capabilities of LLMs is increasingly recognized. Nonetheless, the Korean NLP community lacks publicly available benchmark datasets demonstrating the faithfulness of knowledge-based information. Furthermore, the few existing datasets that evaluate hallucination are limited in their access to the entire dataset, restricting detailed analysis beyond simple scoring, and are based on translated English knowledge. To address these challenges, we introduce K-HALU, a Korean benchmark designed to evaluate LLMs' hallucination detection in Korean. This benchmark contains seven domains, considering the faithfulness of statements based on knowledge documents compiled from Korean news, magazines, and books. For more strict evaluation, 40% of the dataset is structured as multiple-answer questions, requiring models to select all possible correct answers from the given options. Our empirical results show that open-source LLMs still struggle with hallucination detection in Korean knowledge, emphasizing the need for a more detailed analysis of their limitations.

Chaofan Li, Minghao Qin, Shitao Xiao, Jianlyu Chen, Kun Luo, Defu Lian, Yingxia Shao, Zheng Liu

Large language models (LLMs) with decoder-only architectures have demonstrated exceptional text-generation capabilities across a variety of tasks. Some researchers have also adapted these models for text representation tasks. However, in text representation tasks, these models often face performance degradation on unseen tasks. In-context learning (ICL), which leverages examples provided in the input context, enables LLMs to handle unseen tasks effectively. Inspired by this, we aim to fully utilize the inherent properties of LLMs to enhance text representation performance across different tasks through the ICL approach.In this paper, we introduce a simple yet effective training strategy, which significantly improves text representation capabilities. Unlike previous models that prepend task instructions to the text, our method randomly samples a varying number of examples during training, endowing the embedding model with in-context learning abilities while maintaining its zero-shot capabilities. This approach does not require additional data construction or modifications to the model architecture. On the contrary, we find that some popular modifications to the model, such as bidirectional attention, can degrade performance, undermining the inherent characteristics of LLMs. We have publicly released our method at this \href{https://github.com/FlagOpen/FlagEmbedding}{repo}.

Deep Learning · Everything Else

Xingrun Xing, Boyan Gao, Zheng Liu, David Clifton, Shitao Xiao, Wanpeng Zhang, Li Du, Zheng Zhang, Guoqi Li, Jiajun Zhang

Recent advancements in large language models (LLMs) with billions of parameters have improved performance in various applications, but their inference processes demand significant energy and computational resources. In contrast, the human brain, with approximately 86 billion neurons, is much more energy-efficient than LLMs with similar parameters. Inspired by this, we redesign 7$\sim$70 billion parameter LLMs using bio-plausible spiking mechanisms, emulating the efficient behavior of the human brain. We propose the first spiking large language model, SpikeLLM. Coupled with the proposed model, two essential approaches are proposed to improve spike training efficiency: Generalized Integrate-and-Fire (GIF) neurons to compress spike length from $T$ to $\frac{T}{L} \log_2 L$ bits, and an Optimal Brain Spiking framework to divide outlier channels and allocate different $T$ for GIF neurons, which further compresses spike length to approximate $log_2T$ bits. The necessity of spike-driven LLM is proved by comparison with quantized LLMs with similar operations. In the OmniQuant pipeline, SpikeLLM reduces 11.01\% WikiText2 perplexity and improves 2.55\% accuracy of common scene reasoning on a LLAMA-7B W4A4 model. In the GPTQ pipeline, SpikeLLM achieves direct additive in linear layers, significantly exceeding PB-LLMs. Our code is publicly available at https://github.com/Xingrun-Xing2/SpikeLLM.