Large language models (LLMs) have demonstrated remarkable performance across a wide range of natural language processing (NLP) tasks. However, these models are often difficult to deploy due to significant computational requirements and resource constraints. Knowledge distillation (KD) is an effective technique for transferring the performance of larger LLMs to smaller models. Traditional KD methods primarily focus on the direct output of the teacher model, with little emphasis on the role of prompting during knowledge transfer. In this paper, we propose a set of novel response-priming prompting strategies applied in the knowledge distillation pipeline to enhance the performance of student models. Our approach fine-tunes a smaller Llama 3.1 8B Instruct model by distilling knowledge from a quantized Llama 3.1 405B Instruct teacher model. We apply LoRA optimization and evaluate on the GSM8K benchmark. Experimental results demonstrate that integrating reasoning-eliciting prompting into the proposed KD pipeline significantly improves student model performance, offering an efficient way to deploy powerful models in resource-constrained environments. We find that Ground Truth prompting results in a 55\% performance increase on GSM8K for a distilled Llama 3.1 8B Instruct compared to the same model distilled without prompting. A thorough investigation into the self-attention layers of the student models indicates that the more successful prompted models tend to exhibit certain positive behaviors inside their attention heads which can be tied to their increased accuracy. Our implementation can be found at https://github.com/alonso130r/knowledge-distillation.
The takeaway recommendation system is designed to recommend users' future takeaway purchases based on their historical purchase behaviors, thereby improving user satisfaction and increasing merchant sales. Existing methods focus on incorporating auxiliary information or leveraging knowledge graphs to alleviate the sparsity issue of user purchase sequence data. However, two main challenges limit the performance of these approaches: (1) how to capture dynamic user preferences on complex geospatial information and (2) how to efficiently integrate spatial-temporal knowledge from graphs and sequence data with low calculation costs. In this paper, we propose a novel spatial-temporal knowledge distillation for takeaway recommendation model (STKDRec) based on the two-stage training process. Specifically, during the first pre-training stage, a spatial-temporal knowledge graph (STKG) encoder is pre-trained to extract the high-order spatial-temporal and collaborative associations within the STKG. During the second STKD stage, a spatial-temporal Transformer is employed to comprehensively model dynamic user preferences on various types of fine-grained geospatial information from a sequence perspective. Furthermore, the STKD strategy is introduced to adaptively fuse the rich spatial-temporal knowledge from the pre-trained STKG encoder and the spatial-temporal transformer while reducing the cost of model training. Extensive experiments on three real-world datasets show that our STKDRec significantly outperforms the state-of-the-art baselines. Our code is available at:https://github.com/Zhaoshuyuan0246/STKDRec.
Knowledge distillation (KD) is an established paradigm for transferring privileged knowledge from a cumbersome model to a lightweight and efficient one. In recent years, logit-based KD methods are quickly catching up in performance with their feature-based counterparts. However, previous research has pointed out that logit-based methods are still fundamentally limited by two major issues in their training process, namely overconfident teacher and confirmation bias. Inspired by the success of cross-view learning in fields such as semi-supervised learning, in this work we introduce within-view and cross-view regularisations to standard logit-based distillation frameworks to combat the above cruxes. We also perform confidence-based soft label mining to improve the quality of distilling signals from the teacher, which further mitigates the confirmation bias problem. Despite its apparent simplicity, the proposed Consistency-Regularisation-based Logit Distillation (CRLD) significantly boosts student learning, setting new state-of-the-art results on the standard CIFAR-100, Tiny-ImageNet, and ImageNet datasets across a diversity of teacher and student architectures, whilst introducing no extra network parameters. Orthogonal to on-going logit-based distillation research, our method enjoys excellent generalisation properties and, without bells and whistles, boosts the performance of various existing approaches by considerable margins.
Recently, radar-camera fusion algorithms have gained significant attention as radar sensors provide geometric information that complements the limitations of cameras. However, most existing radar-camera depth estimation algorithms focus solely on improving performance, often neglecting computational efficiency. To address this gap, we propose LiRCDepth, a lightweight radar-camera depth estimation model. We incorporate knowledge distillation to enhance the training process, transferring critical information from a complex teacher model to our lightweight student model in three key domains. Firstly, low-level and high-level features are transferred by incorporating pixel-wise and pair-wise distillation. Additionally, we introduce an uncertainty-aware inter-depth distillation loss to refine intermediate depth maps during decoding. Leveraging our proposed knowledge distillation scheme, the lightweight model achieves a 6.6% improvement in MAE on the nuScenes dataset compared to the model trained without distillation. Code: https://github.com/harborsarah/LiRCDepth
Recent advances in diffusion generative models have yielded remarkable progress. While the quality of generated content continues to improve, these models have grown considerably in size and complexity. This increasing computational burden poses significant challenges, particularly in resource-constrained deployment scenarios such as mobile devices. The combination of model pruning and knowledge distillation has emerged as a promising solution to reduce computational demands while preserving generation quality. However, this technique inadvertently propagates undesirable behaviors, including the generation of copyrighted content and unsafe concepts, even when such instances are absent from the fine-tuning dataset. In this paper, we propose a novel bilevel optimization framework for pruned diffusion models that consolidates the fine-tuning and unlearning processes into a unified phase. Our approach maintains the principal advantages of distillation-namely, efficient convergence and style transfer capabilities-while selectively suppressing the generation of unwanted content. This plug-in framework is compatible with various pruning and concept unlearning methods, facilitating efficient, safe deployment of diffusion models in controlled environments.
Knowledge distillation (KD) has become a prevalent technique for compressing large language models (LLMs). Existing KD methods are constrained by the need for identical tokenizers (i.e., vocabularies) between teacher and student models, limiting their versatility in handling LLMs of different architecture families. In this paper, we introduce the Multi-Level Optimal Transport (MultiLevelOT), a novel approach that advances the optimal transport for universal cross-tokenizer knowledge distillation. Our method aligns the logit distributions of the teacher and the student at both token and sequence levels using diverse cost matrices, eliminating the need for dimensional or token-by-token correspondence. At the token level, MultiLevelOT integrates both global and local information by jointly optimizing all tokens within a sequence to enhance robustness. At the sequence level, we efficiently capture complex distribution structures of logits via the Sinkhorn distance, which approximates the Wasserstein distance for divergence measures. Extensive experiments on tasks such as extractive QA, generative QA, and summarization demonstrate that the MultiLevelOT outperforms state-of-the-art cross-tokenizer KD methods under various settings. Our approach is robust to different student and teacher models across model families, architectures, and parameter sizes. Codes and models are available at https://github.com/2018cx/Multi-Level-OT.
Educational data mining (EDM) is a part of applied computing that focuses on automatically analyzing data from learning contexts. Early prediction for identifying at-risk students is a crucial and widely researched topic in EDM research. It enables instructors to support at-risk students to stay on track, preventing student dropout or failure. Previous studies have predicted students' learning performance to identify at-risk students by using machine learning on data collected from e-learning platforms. However, most studies aimed to identify at-risk students utilizing the entire course data after the course finished. This does not correspond to the real-world scenario that at-risk students may drop out before the course ends. To address this problem, we introduce an RNN-Attention-KD (knowledge distillation) framework to predict at-risk students early throughout a course. It leverages the strengths of Recurrent Neural Networks (RNNs) in handling time-sequence data to predict students' performance at each time step and employs an attention mechanism to focus on relevant time steps for improved predictive accuracy. At the same time, KD is applied to compress the time steps to facilitate early prediction. In an empirical evaluation, RNN-Attention-KD outperforms traditional neural network models in terms of recall and F1-measure. For example, it obtained recall and F1-measure of 0.49 and 0.51 for Weeks 1--3 and 0.51 and 0.61 for Weeks 1--6 across all datasets from four years of a university course. Then, an ablation study investigated the contributions of different knowledge transfer methods (distillation objectives). We found that hint loss from the hidden layer of RNN and context vector loss from the attention module on RNN could enhance the model's prediction performance for identifying at-risk students. These results are relevant for EDM researchers employing deep learning models.
The emergence of large-scale Mixture of Experts (MoE) models represents a significant advancement in artificial intelligence, offering enhanced model capacity and computational efficiency through conditional computation. However, deploying and running inference on these models presents significant challenges in computational resources, latency, and energy efficiency. This comprehensive survey analyzes optimization techniques for MoE models across the entire system stack. We first establish a taxonomical framework that categorizes optimization approaches into model-level, system-level, and hardware-level optimizations. At the model level, we examine architectural innovations including efficient expert design, attention mechanisms, various compression techniques such as pruning, quantization, and knowledge distillation, as well as algorithm improvement including dynamic routing strategies and expert merging methods. At the system level, we investigate distributed computing approaches, load balancing mechanisms, and efficient scheduling algorithms that enable scalable deployment. Furthermore, we delve into hardware-specific optimizations and co-design strategies that maximize throughput and energy efficiency. This survey provides both a structured overview of existing solutions and identifies key challenges and promising research directions in MoE inference optimization. To facilitate ongoing updates and the sharing of cutting-edge advances in MoE inference optimization research, we have established a repository accessible at https://github.com/MoE-Inf/awesome-moe-inference/.
The limited availability of labeled data has driven advancements in semi-supervised learning for medical image segmentation. Modern large-scale models tailored for general segmentation, such as the Segment Anything Model (SAM), have revealed robust generalization capabilities. However, applying these models directly to medical image segmentation still exposes performance degradation. In this paper, we propose a learnable prompting SAM-induced Knowledge distillation framework (KnowSAM) for semi-supervised medical image segmentation. Firstly, we propose a Multi-view Co-training (MC) strategy that employs two distinct sub-networks to employ a co-teaching paradigm, resulting in more robust outcomes. Secondly, we present a Learnable Prompt Strategy (LPS) to dynamically produce dense prompts and integrate an adapter to fine-tune SAM specifically for medical image segmentation tasks. Moreover, we propose SAM-induced Knowledge Distillation (SKD) to transfer useful knowledge from SAM to two sub-networks, enabling them to learn from SAM's predictions and alleviate the effects of incorrect pseudo-labels during training. Notably, the predictions generated by our subnets are used to produce mask prompts for SAM, facilitating effective inter-module information exchange. Extensive experimental results on various medical segmentation tasks demonstrate that our model outperforms the state-of-the-art semi-supervised segmentation approaches. Crucially, our SAM distillation framework can be seamlessly integrated into other semi-supervised segmentation methods to enhance performance. The code will be released upon acceptance of this manuscript at: https://github.com/taozh2017/KnowSAM
Data-free knowledge distillation aims to learn a compact student network from a pre-trained large teacher network without using the original training data of the teacher network. Existing collection-based and generation-based methods train student networks by collecting massive real examples and generating synthetic examples, respectively. However, they inevitably become weak in practical scenarios due to the difficulties in gathering or emulating sufficient real-world data. To solve this problem, we propose a novel method called \textbf{H}ybr\textbf{i}d \textbf{D}ata-\textbf{F}ree \textbf{D}istillation (HiDFD), which leverages only a small amount of collected data as well as generates sufficient examples for training student networks. Our HiDFD comprises two primary modules, \textit{i.e.}, the teacher-guided generation and student distillation. The teacher-guided generation module guides a Generative Adversarial Network (GAN) by the teacher network to produce high-quality synthetic examples from very few real-world collected examples. Specifically, we design a feature integration mechanism to prevent the GAN from overfitting and facilitate the reliable representation learning from the teacher network. Meanwhile, we drive a category frequency smoothing technique via the teacher network to balance the generative training of each category. In the student distillation module, we explore a data inflation strategy to properly utilize a blend of real and synthetic data to train the student network via a classifier-sharing-based feature alignment technique. Intensive experiments across multiple benchmarks demonstrate that our HiDFD can achieve state-of-the-art performance using 120 times less collected data than existing methods. Code is available at https://github.com/tangjialiang97/HiDFD.
We highlight two significant issues leading to the passivity of current merchant non-player characters (NPCs): pricing and communication. While immersive interactions have been a focus, negotiations between merchant NPCs and players on item prices have not received sufficient attention. First, we define passive pricing as the limited ability of merchants to modify predefined item prices. Second, passive communication means that merchants can only interact with players in a scripted manner. To tackle these issues and create an active merchant NPC, we propose a merchant framework based on large language models (LLMs), called MART, which consists of an appraiser module and a negotiator module. We conducted two experiments to guide game developers in selecting appropriate implementations by comparing different training methods and LLM sizes. Our findings indicate that finetuning methods, such as supervised finetuning (SFT) and knowledge distillation (KD), are effective in using smaller LLMs to implement active merchant NPCs. Additionally, we found three irregular cases arising from the responses of LLMs. We expect our findings to guide developers in using LLMs for developing active merchant NPCs.
In this paper, we propose $\textbf{Ne}$ural-$\textbf{Sy}$mbolic $\textbf{C}$ollaborative $\textbf{D}$istillation ($\textbf{NesyCD}$), a novel knowledge distillation method for learning the complex reasoning abilities of Large Language Models (LLMs, e.g., \textgreater 13B). We argue that complex reasoning tasks are difficult for Small Language Models (SLMs, e.g., $\leq$ 7B), as these tasks demand not only general cognitive abilities but also specialized knowledge, which is often sparse and difficult for these neural-based SLMs to effectively capture. Therefore, NesyCD distills the general capabilities and specialized knowledge in LLMs using different manners. On the one hand, we distill only general abilities from teacher LLMs into the student SLMs of parameterized neural networks. On the other hand, for the specialized abilities and uncommon knowledge of a complex reasoning task, we employ a symbolic knowledge distillation approach to obtain and store the specialized knowledge within a symbolic knowledge base (KB). By decoupling general and specialized capabilities, the proposed NesyCD can achieve superior performance cost-effectively, utilizing smaller models and blending parameterized neural networks with symbolic KB. Moreover, the specialized KB generalizes well and is comprehended and manipulated by humans. Our experiments show that NesyCD significantly boosts SLMs' complex reasoning performance on in-domain (BBH, GSM8K) and out-of-domain (AGIEval, ARC) datasets. Notably, our approach enabled the LLaMA3-8B and Qwen2-7B to surpass GPT-3.5-turbo in performance and come close to matching LLaMA3-70B, despite the latter having nine times more parameters. Our code will be available at https://github.com/Xnhyacinth/NesyCD.
Data-free knowledge distillation transfers knowledge by recovering training data from a pre-trained model. Despite the recent success of seeking global data diversity, the diversity within each class and the similarity among different classes are largely overlooked, resulting in data homogeneity and limited performance. In this paper, we introduce a novel Relation-Guided Adversarial Learning method with triplet losses, which solves the homogeneity problem from two aspects. To be specific, our method aims to promote both intra-class diversity and inter-class confusion of the generated samples. To this end, we design two phases, an image synthesis phase and a student training phase. In the image synthesis phase, we construct an optimization process to push away samples with the same labels and pull close samples with different labels, leading to intra-class diversity and inter-class confusion, respectively. Then, in the student training phase, we perform an opposite optimization, which adversarially attempts to reduce the distance of samples of the same classes and enlarge the distance of samples of different classes. To mitigate the conflict of seeking high global diversity and keeping inter-class confusing, we propose a focal weighted sampling strategy by selecting the negative in the triplets unevenly within a finite range of distance. RGAL shows significant improvement over previous state-of-the-art methods in accuracy and data efficiency. Besides, RGAL can be inserted into state-of-the-art methods on various data-free knowledge transfer applications. Experiments on various benchmarks demonstrate the effectiveness and generalizability of our proposed method on various tasks, specially data-free knowledge distillation, data-free quantization, and non-exemplar incremental learning. Our code is available at https://github.com/Sharpiless/RGAL.
In the realm of novelty detection, accurately identifying outliers in data without specific class information poses a significant challenge. While current methods excel in single-object scenarios, they struggle with multi-object situations due to their focus on individual objects. Our paper suggests a novel approach: redefining `normal' at the object level in training datasets. Rather than the usual image-level view, we consider the most dominant object in a dataset as the norm, offering a perspective that is more effective for real-world scenarios. Adapting to our object-level definition of `normal', we modify knowledge distillation frameworks, where a student network learns from a pre-trained teacher network. Our first contribution, DeFeND(Dense Feature Fine-tuning on Normal Data), integrates dense feature fine-tuning into the distillation process, allowing the teacher network to focus on object-level features with a self-supervised loss. The second is masked knowledge distillation, where the student network works with partially hidden inputs, honing its ability to deduce and generalize from incomplete data. This approach not only fares well in single-object novelty detection but also considerably surpasses existing methods in multi-object contexts. The implementation is available at: https://github.com/SMSD75/Redefining_Normal_ACCV24/tree/main
Knowledge distillation (KD) has been widely studied in unsupervised Industrial Image Anomaly Detection (AD), but its application to unsupervised multimodal AD remains underexplored. Existing KD-based methods for multimodal AD that use fused multimodal features to obtain teacher representations face challenges. Anomalies in one modality may not be effectively captured in the fused teacher features, leading to detection failures. Besides, these methods do not fully leverage the rich intra- and inter-modality information. In this paper, we propose Crossmodal Reverse Distillation (CRD) based on Multi-branch design to realize Multimodal Industrial AD. By assigning independent branches to each modality, our method enables finer detection of anomalies within each modality. Furthermore, we enhance the interaction between modalities during the distillation process by designing Crossmodal Filter and Amplifier. With the idea of crossmodal mapping, the student network is allowed to better learn normal features while anomalies in all modalities are ensured to be effectively detected. Experimental verifications on the MVTec 3D-AD dataset demonstrate that our method achieves state-of-the-art performance in multimodal anomaly detection and localization.
Knowledge distillation (KD) is a valuable yet challenging approach that enhances a compact student network by learning from a high-performance but cumbersome teacher model. However, previous KD methods for image restoration overlook the state of the student during the distillation, adopting a fixed solution space that limits the capability of KD. Additionally, relying solely on L1-type loss struggles to leverage the distribution information of images. In this work, we propose a novel dynamic contrastive knowledge distillation (DCKD) framework for image restoration. Specifically, we introduce dynamic contrastive regularization to perceive the student's learning state and dynamically adjust the distilled solution space using contrastive learning. Additionally, we also propose a distribution mapping module to extract and align the pixel-level category distribution of the teacher and student models. Note that the proposed DCKD is a structure-agnostic distillation framework, which can adapt to different backbones and can be combined with methods that optimize upper-bound constraints to further enhance model performance. Extensive experiments demonstrate that DCKD significantly outperforms the state-of-the-art KD methods across various image restoration tasks and backbones.
Lifelong person re-identification (LReID) is an important but challenging task that suffers from catastrophic forgetting due to significant domain gaps between training steps. Existing LReID approaches typically rely on data replay and knowledge distillation to mitigate this issue. However, data replay methods compromise data privacy by storing historical exemplars, while knowledge distillation methods suffer from limited performance due to the cumulative forgetting of undistilled knowledge. To overcome these challenges, we propose a novel paradigm that models and rehearses the distribution of the old domains to enhance knowledge consolidation during the new data learning, possessing a strong anti-forgetting capacity without storing any exemplars. Specifically, we introduce an exemplar-free LReID method called Distribution Rehearsing via Adaptive Style Kernel Learning (DASK). DASK includes a Distribution Rehearser Learning (DRL) mechanism that learns to transform arbitrary distribution data into the current data style at each learning step. To enhance the style transfer capacity of DRL, an Adaptive Kernel Prediction Network (AKPNet) is explored to achieve an instance-specific distribution adjustment. Additionally, we design a Distribution Rehearsing-driven LReID Training (DRRT) module, which rehearses old distribution based on the new data via the old AKPNet model, achieving effective new-old knowledge accumulation under a joint knowledge consolidation scheme. Experimental results show our DASK outperforms the existing methods by 3.6%-6.8% and 4.5%-6.5% on anti-forgetting and generalization capacity, respectively. Our code is available at https://github.com/zhoujiahuan1991/AAAI2025-LReID-DASK
Knowledge Distillation (KD) is a promising approach for unsupervised Anomaly Detection (AD). However, the student network's over-generalization often diminishes the crucial representation differences between teacher and student in anomalous regions, leading to detection failures. To addresses this problem, the widely accepted Reverse Distillation (RD) paradigm designs the asymmetry teacher and student, using an encoder as teacher and a decoder as student. Yet, the design of RD does not ensure that the teacher encoder effectively distinguishes between normal and abnormal features or that the student decoder generates anomaly-free features. Additionally, the absence of skip connections results in a loss of fine details during feature reconstruction. To address these issues, we propose RD with Expert, which introduces a novel Expert-Teacher-Student network for simultaneous distillation of both the teacher encoder and student decoder. The added expert network enhances the student's ability to generate normal features and optimizes the teacher's differentiation between normal and abnormal features, reducing missed detections. Additionally, Guided Information Injection is designed to filter and transfer features from teacher to student, improving detail reconstruction and minimizing false positives. Experiments on several benchmarks prove that our method outperforms existing unsupervised AD methods under RD paradigm, fully unlocking RD's potential.
Multimodal multihop question answering (MMQA) requires reasoning over images and text from multiple sources. Despite advances in visual question answering, this multihop setting remains underexplored due to a lack of quality datasets. Existing methods focus on single-hop, single-modality, or short texts, limiting real-world applications like interpreting educational documents with long, multimodal content. To fill this gap, we introduce FM2DS, the first framework for creating a high-quality dataset for MMQA. Our approach consists of a 5-stage pipeline that involves acquiring relevant multimodal documents from Wikipedia, synthetically generating high-level questions and answers, and validating them through rigorous criteria to ensure data quality. We evaluate our methodology by training models on our synthesized dataset and testing on two benchmarks: MultimodalQA and WebQA. Our results demonstrate that, with an equal sample size, models trained on our synthesized data outperform those trained on human-collected data by 1.9 in exact match (EM) score on average. Additionally, we introduce M2QA-Bench with 1k samples, the first benchmark for MMQA on long documents, generated using FM2DS and refined by human annotators. We believe our data synthesis method will serve as a strong foundation for training and evaluating MMQA models.
Federated Learning (FL) is an innovative distributed machine learning paradigm that enables neural network training across devices without centralizing data. While this addresses issues of information sharing and data privacy, challenges arise from data heterogeneity across clients and increasing network scale, leading to impacts on model performance and training efficiency. Previous research shows that in IID environments, the parameter structure of the model is expected to adhere to certain specific consistency principles. Thus, identifying and regularizing these consistencies can mitigate issues from heterogeneous data. We found that both soft labels derived from knowledge distillation and the classifier head parameter matrix, when multiplied by their own transpose, capture the intrinsic relationships between data classes. These shared relationships suggest inherent consistency. Therefore, the work in this paper identifies the consistency between the two and leverages it to regulate training, underpinning our proposed FedDW framework. Experimental results show FedDW outperforms 10 state-of-the-art FL methods, improving accuracy by an average of 3% in highly heterogeneous settings. Additionally, we provide a theoretical proof that FedDW offers higher efficiency, with the additional computational load from backpropagation being negligible. The code is available at https://github.com/liuvvvvv1/FedDW.