Search

Ensuring the safety of the Large Language Model (LLM) is critical, but currently used methods in most cases sacrifice the model performance to obtain increased safety or perform poorly on data outside of their adaptation distribution. We investigate existing methods for such generalization and find them insufficient. Surprisingly, while even plain LLMs recognize unsafe prompts, they may still generate unsafe responses. To avoid performance degradation and preserve safe performance, we advocate for a two-step framework, where we first identify unsafe prompts via a lightweight classifier, and apply a "safe" model only to such prompts. In particular, we explore the design of the safety detector in more detail, investigating the use of different classifier architectures and prompting techniques. Interestingly, we find that the final hidden state for the last token is enough to provide robust performance, minimizing false positives on benign data while performing well on malicious prompt detection. Additionally, we show that classifiers trained on the representations from different model layers perform comparably on the latest model layers, indicating that safety representation is present in the LLMs' hidden states at most model stages. Our work is a step towards efficient, representation-based safety mechanisms for LLMs.

Continual learning (CL) has emerged as a critical area in machine learning, enabling neural networks to learn from evolving data distributions while mitigating catastrophic forgetting. However, recent research has identified the stability gap -- a phenomenon where models initially lose performance on previously learned tasks before partially recovering during training. Such learning dynamics are contradictory to the intuitive understanding of stability in continual learning where one would expect the performance to degrade gradually instead of rapidly decreasing and then partially recovering later. To better understand and alleviate the stability gap, we investigate it at different levels of the neural network architecture, particularly focusing on the role of the classification head. We introduce the nearest-mean classifier (NMC) as a tool to attribute the influence of the backbone and the classification head on the stability gap. Our experiments demonstrate that NMC not only improves final performance, but also significantly enhances training stability across various continual learning benchmarks, including CIFAR100, ImageNet100, CUB-200, and FGVC Aircrafts. Moreover, we find that NMC also reduces task-recency bias. Our analysis provides new insights into the stability gap and suggests that the primary contributor to this phenomenon is the linear head, rather than the insufficient representation learning., Comment: Accepted at WACV 2025

Static sparse training aims to train sparse models from scratch, achieving remarkable results in recent years. A key design choice is given by the sparse initialization, which determines the trainable sub-network through a binary mask. Existing methods mainly select such mask based on a predefined dense initialization. Such an approach may not efficiently leverage the mask's potential impact on the optimization. An alternative direction, inspired by research into dynamical isometry, is to introduce orthogonality in the sparse subnetwork, which helps in stabilizing the gradient signal. In this work, we propose Exact Orthogonal Initialization (EOI), a novel sparse orthogonal initialization scheme based on composing random Givens rotations. Contrary to other existing approaches, our method provides exact (not approximated) orthogonality and enables the creation of layers with arbitrary densities. We demonstrate the superior effectiveness and efficiency of EOI through experiments, consistently outperforming common sparse initialization techniques. Our method enables training highly sparse 1000-layer MLP and CNN networks without residual connections or normalization techniques, emphasizing the crucial role of weight initialization in static sparse training alongside sparse mask selection. The code is available at https://github.com/woocash2/sparser-better-deeper-stronger, Comment: ICML 2024

Continual learning is crucial for applications in dynamic environments, where machine learning models must adapt to changing data distributions while retaining knowledge of previous tasks. Despite significant advancements, catastrophic forgetting - where performance on earlier tasks degrades as new information is learned - remains a key challenge. In this work, we investigate the stability of intermediate neural network layers during continual learning and explore how auxiliary classifiers (ACs) can leverage this stability to improve performance. We show that early network layers remain more stable during learning, particularly for older tasks, and that ACs applied to these layers can outperform standard classifiers on past tasks. By integrating ACs into several continual learning algorithms, we demonstrate consistent and significant performance improvements on standard benchmarks. Additionally, we explore dynamic inference, showing that AC-augmented continual learning methods can reduce computational costs by up to 60\% while maintaining or exceeding the accuracy of standard methods. Our findings suggest that ACs offer a promising avenue for enhancing continual learning models, providing both improved performance and the ability to adapt the network computation in environments where such flexibility might be required.

Transformer models can face practical limitations due to their high computational requirements. At the same time, such models exhibit significant activation sparsity, which can be leveraged to reduce the inference cost by converting parts of the network into equivalent Mixture-of-Experts (MoE) layers. Despite the crucial role played by activation sparsity, its impact on this process remains unexplored. We demonstrate that the efficiency of the conversion can be significantly enhanced by a proper regularization of the activation sparsity of the base model. Moreover, motivated by the high variance of the number of activated neurons for different inputs, we introduce a more effective dynamic-$k$ expert selection rule that adjusts the number of executed experts on a per-token basis. To achieve further savings, we extend this approach to multi-head attention projections. Finally, we develop an efficient implementation that translates these computational savings into actual wall-clock speedup. The proposed method, Dense to Dynamic-$k$ Mixture-of-Experts (D2DMoE), outperforms existing approaches on common NLP and vision tasks, reducing inference cost by up to 60% without significantly impacting performance.

In this work, we investigate exemplar-free class incremental learning (CIL) with knowledge distillation (KD) as a regularization strategy, aiming to prevent forgetting. KD-based methods are successfully used in CIL, but they often struggle to regularize the model without access to exemplars of the training data from previous tasks. Our analysis reveals that this issue originates from substantial representation shifts in the teacher network when dealing with out-of-distribution data. This causes large errors in the KD loss component, leading to performance degradation in CIL models. Inspired by recent test-time adaptation methods, we introduce Teacher Adaptation (TA), a method that concurrently updates the teacher and the main models during incremental training. Our method seamlessly integrates with KD-based CIL approaches and allows for consistent enhancement of their performance across multiple exemplar-free CIL benchmarks. The source code for our method is available at https://github.com/fszatkowski/cl-teacher-adaptation., Comment: Accepted at WACV 2024

Implicit Neural Representations (INRs) are nowadays used to represent multimedia signals across various real-life applications, including image super-resolution, image compression, or 3D rendering. Existing methods that leverage INRs are predominantly focused on visual data, as their application to other modalities, such as audio, is nontrivial due to the inductive biases present in architectural attributes of image-based INR models. To address this limitation, we introduce HyperSound, the first meta-learning approach to produce INRs for audio samples that leverages hypernetworks to generalize beyond samples observed in training. Our approach reconstructs audio samples with quality comparable to other state-of-the-art models and provides a viable alternative to contemporary sound representations used in deep neural networks for audio processing, such as spectrograms., Comment: ECML2023

Implicit neural representations (INRs) are a rapidly growing research field, which provides alternative ways to represent multimedia signals. Recent applications of INRs include image super-resolution, compression of high-dimensional signals, or 3D rendering. However, these solutions usually focus on visual data, and adapting them to the audio domain is not trivial. Moreover, it requires a separately trained model for every data sample. To address this limitation, we propose HyperSound, a meta-learning method leveraging hypernetworks to produce INRs for audio signals unseen at training time. We show that our approach can reconstruct sound waves with quality comparable to other state-of-the-art models., Comment: NeurIPS 2022 MetaLearn workshop

We introduce a new method for internal replay that modulates the frequency of rehearsal based on the depth of the network. While replay strategies mitigate the effects of catastrophic forgetting in neural networks, recent works on generative replay show that performing the rehearsal only on the deeper layers of the network improves the performance in continual learning. However, the generative approach introduces additional computational overhead, limiting its applications. Motivated by the observation that earlier layers of neural networks forget less abruptly, we propose to update network layers with varying frequency using intermediate-level features during replay. This reduces the computational burden by omitting computations for both deeper layers of the generator and earlier layers of the main model. We name our method Progressive Latent Replay and show that it outperforms Internal Replay while using significantly fewer resources.