arXiv subjects

Axel Laborieux

Publications and source records attributed to Axel Laborieux.

16 recordsLinked to original sources

arXiv2026

Blurry Window Attention

The Softmax Attention operation in Transformer language models has a quadratic complexity in the sequence length and a growing state size in the form of KV cache, which becomes a bottleneck in long context scenarios. To overcome this limitation, alternative architectures with linear complexity and finite state size have been introduced, such as State-Space Models (SSMs), Linear Attention (LA), and Attention with Bounded-memory Control (ABC). Though linear models achieve similar language perplexity as Transformers, they are still behind in tasks which require retrieval or recall of specific information. In this work, we introduce Blurry Window Attention (BLA) a novel ABC method inspired by SSMs. BLA stores a frequency window from which a blurry KV history is reconstructed via interpolation using Dirichlet kernels. BLA can be understood as a generalization of Sliding Window Attention (SWA) depending on the Dirichlet kernels resolution or as a special case of the Gated Slot Attention (GSA), where the decay factor is implemented with Dirichlet kernels. We describe in details the theory and efficient implementation of BLA. On the Multi-Query Associate Recall (MQAR) synthetic task, we show that the state efficiency of BLA is 8$\times$ better than SWA and is competitive with popular linear attention models, and in the RegBench synthetic task, only BLA and SWA improve their performance as the state size grows among the linear models we tested.

cs.LG↗

arXiv2026

AllMem: A Memory-centric Recipe for Efficient Long-context Modeling

Large Language Models (LLMs) encounter significant performance bottlenecks in long-sequence tasks due to the computational complexity and memory overhead inherent in the self-attention mechanism. To address these challenges, we introduce \textsc{AllMem}, a novel and efficient hybrid architecture that integrates Sliding Window Attention (SWA) with non-linear Test-Time Training (TTT) memory networks. \textsc{AllMem} enables models to effectively scale to ultra-long contexts while mitigating catastrophic forgetting. This approach not only overcomes the representation constraints typical of linear memory models but also significantly reduces the computational and memory footprint during long-sequence inference. Furthermore, we implement a Memory-Efficient Fine-Tuning strategy to replace standard attention layers in pre-trained models with memory-augmented sliding window layers. This framework facilitates the efficient transformation of any off-the-shelf pre-trained LLM into an \textsc{AllMem}-based architecture. Empirical evaluations confirm that our 4k window model achieves near-lossless performance on 37k LongBench with a marginal 0.83 drop compared to full attention. Furthermore, on InfiniteBench at a 128k context, our 8k window variant outperforms full attention, which validates the effectiveness of our parameterized memory in mitigating noise and maintaining robust long-range modeling without the prohibitive costs of global attention.

cs.AI↗

arXiv2025

Teaching signal synchronization in deep neural networks with prospective neurons

Working memory requires the brain to maintain information from the recent past to guide ongoing behavior. Neurons can contribute to this capacity by slowly integrating their inputs over time, creating persistent activity that outlasts the original stimulus. However, when these slowly integrating neurons are organized hierarchically, they introduce cumulative delays that create a fundamental challenge for learning: teaching signals that indicate whether behavior was correct or incorrect arrive out-of-sync with the neural activity they are meant to instruct. Here, we demonstrate that neurons enhanced with an adaptive current can compensate for these delays by responding to external stimuli prospectively -- effectively predicting future inputs to synchronize with them. First, we show that such prospective neurons enable teaching signal synchronization across a range of learning algorithms that propagate error signals through hierarchical networks. Second, we demonstrate that this successfully guides learning in slowly integrating neurons, enabling the formation and retrieval of memories over extended timescales. We support our findings with a mathematical analysis of the prospective coding mechanism and learning experiments on motor control tasks. Together, our results reveal how neural adaptation could solve a critical timing problem and enable efficient learning in dynamic environments.

q-bio.NC↗

arXiv2025

Memory-Augmented Transformers: A Systematic Review from Neuroscience Principles to Enhanced Model Architectures

Memory is fundamental to intelligence, enabling learning, reasoning, and adaptability across biological and artificial systems. While Transformer architectures excel at sequence modeling, they face critical limitations in long-range context retention, continual learning, and knowledge integration. This review presents a unified framework bridging neuroscience principles, including dynamic multi-timescale memory, selective attention, and consolidation, with engineering advances in Memory-Augmented Transformers. We organize recent progress through three taxonomic dimensions: functional objectives (context extension, reasoning, knowledge integration, adaptation), memory representations (parameter-encoded, state-based, explicit, hybrid), and integration mechanisms (attention fusion, gated control, associative retrieval). Our analysis of core memory operations (reading, writing, forgetting, and capacity management) reveals a shift from static caches toward adaptive, test-time learning systems. We identify persistent challenges in scalability and interference, alongside emerging solutions including hierarchical buffering and surprise-gated updates. This synthesis provides a roadmap toward cognitively-inspired, lifelong-learning Transformer architectures.

cs.LG↗

arXiv2024

Theories of synaptic memory consolidation and intelligent plasticity for continual learning

Humans and animals learn throughout life. Such continual learning is crucial for intelligence. In this chapter, we examine the pivotal role plasticity mechanisms with complex internal synaptic dynamics could play in enabling this ability in neural networks. By surveying theoretical research, we highlight two fundamental enablers for continual learning. First, synaptic plasticity mechanisms must maintain and evolve an internal state over several behaviorally relevant timescales. Second, plasticity algorithms must leverage the internal state to intelligently regulate plasticity at individual synapses to facilitate the seamless integration of new memories while avoiding detrimental interference with existing ones. Our chapter covers successful applications of these principles to deep neural networks and underscores the significance of synaptic metaplasticity in sustaining continual learning capabilities. Finally, we outline avenues for further research to understand the brain's superb continual learning abilities and harness similar mechanisms for artificial intelligence systems.

q-bio.NC↗

arXiv2023

Implicit variance regularization in non-contrastive SSL

Non-contrastive SSL methods like BYOL and SimSiam rely on asymmetric predictor networks to avoid representational collapse without negative samples. Yet, how predictor networks facilitate stable learning is not fully understood. While previous theoretical analyses assumed Euclidean losses, most practical implementations rely on cosine similarity. To gain further theoretical insight into non-contrastive SSL, we analytically study learning dynamics in conjunction with Euclidean and cosine similarity in the eigenspace of closed-form linear predictor networks. We show that both avoid collapse through implicit variance regularization albeit through different dynamical mechanisms. Moreover, we find that the eigenvalues act as effective learning rate multipliers and propose a family of isotropic loss functions (IsoLoss) that equalize convergence rates across eigenmodes. Empirically, IsoLoss speeds up the initial learning dynamics and increases robustness, thereby allowing us to dispense with the EMA target network typically used with non-contrastive methods. Our analysis sheds light on the variance regularization mechanisms of non-contrastive SSL and lays the theoretical grounds for crafting novel loss functions that shape the learning dynamics of the predictor's spectrum.

cs.LG↗

arXiv2023

Improving equilibrium propagation without weight symmetry through Jacobian homeostasis

Equilibrium propagation (EP) is a compelling alternative to the backpropagation of error algorithm (BP) for computing gradients of neural networks on biological or analog neuromorphic substrates. Still, the algorithm requires weight symmetry and infinitesimal equilibrium perturbations, i.e., nudges, to estimate unbiased gradients efficiently. Both requirements are challenging to implement in physical systems. Yet, whether and how weight asymmetry affects its applicability is unknown because, in practice, it may be masked by biases introduced through the finite nudge. To address this question, we study generalized EP, which can be formulated without weight symmetry, and analytically isolate the two sources of bias. For complex-differentiable non-symmetric networks, we show that the finite nudge does not pose a problem, as exact derivatives can still be estimated via a Cauchy integral. In contrast, weight asymmetry introduces bias resulting in low task performance due to poor alignment of EP's neuronal error vectors compared to BP. To mitigate this issue, we present a new homeostatic objective that directly penalizes functional asymmetries of the Jacobian at the network's fixed point. This homeostatic objective dramatically improves the network's ability to solve complex tasks such as ImageNet 32x32. Our results lay the theoretical groundwork for studying and mitigating the adverse effects of imperfections of physical networks on learning algorithms that rely on the substrate's relaxation dynamics.

cs.LG↗

arXiv2022

Holomorphic Equilibrium Propagation Computes Exact Gradients Through Finite Size Oscillations

Equilibrium propagation (EP) is an alternative to backpropagation (BP) that allows the training of deep neural networks with local learning rules. It thus provides a compelling framework for training neuromorphic systems and understanding learning in neurobiology. However, EP requires infinitesimal teaching signals, thereby limiting its applicability in noisy physical systems. Moreover, the algorithm requires separate temporal phases and has not been applied to large-scale problems. Here we address these issues by extending EP to holomorphic networks. We show analytically that this extension naturally leads to exact gradients even for finite-amplitude teaching signals. Importantly, the gradient can be computed as the first Fourier coefficient from finite neuronal activity oscillations in continuous time without requiring separate phases. Further, we demonstrate in numerical simulations that our approach permits robust estimation of gradients in the presence of noise and that deeper models benefit from the finite teaching signals. Finally, we establish the first benchmark for EP on the ImageNet 32x32 dataset and show that it matches the performance of an equivalent network trained with BP. Our work provides analytical insights that enable scaling EP to large-scale problems and establishes a formal framework for how oscillations could support learning in biological and neuromorphic systems.

cs.LG↗

arXiv2021

Model of the Weak Reset Process in HfOx Resistive Memory for Deep Learning Frameworks

The implementation of current deep learning training algorithms is power-hungry, owing to data transfer between memory and logic units. Oxide-based RRAMs are outstanding candidates to implement in-memory computing, which is less power-intensive. Their weak RESET regime, is particularly attractive for learning, as it allows tuning the resistance of the devices with remarkable endurance. However, the resistive change behavior in this regime suffers many fluctuations and is particularly challenging to model, especially in a way compatible with tools used for simulating deep learning. In this work, we present a model of the weak RESET process in hafnium oxide RRAM and integrate this model within the PyTorch deep learning framework. Validated on experiments on a hybrid CMOS/RRAM technology, our model reproduces both the noisy progressive behavior and the device-to-device (D2D) variability. We use this tool to train Binarized Neural Networks for the MNIST handwritten digit recognition task and the CIFAR-10 object classification task. We simulate our model with and without various aspects of device imperfections to understand their impact on the training process and identify that the D2D variability is the most detrimental aspect. The framework can be used in the same manner for other types of memories to identify the device imperfections that cause the most degradation, which can, in turn, be used to optimize the devices to reduce the impact of these imperfections.

cs.LG↗

arXiv2021

Synaptic Metaplasticity in Binarized Neural Networks

While deep neural networks have surpassed human performance in multiple situations, they are prone to catastrophic forgetting: upon training a new task, they rapidly forget previously learned ones. Neuroscience studies, based on idealized tasks, suggest that in the brain, synapses overcome this issue by adjusting their plasticity depending on their past history. However, such "metaplastic" behaviours do not transfer directly to mitigate catastrophic forgetting in deep neural networks. In this work, we interpret the hidden weights used by binarized neural networks, a low-precision version of deep neural networks, as metaplastic variables, and modify their training technique to alleviate forgetting. Building on this idea, we propose and demonstrate experimentally, in situations of multitask and stream learning, a training technique that reduces catastrophic forgetting without needing previously presented data, nor formal boundaries between datasets and with performance approaching more mainstream techniques with task boundaries. We support our approach with a theoretical analysis on a tractable task. This work bridges computational neuroscience and deep learning, and presents significant assets for future embedded and neuromorphic systems, especially when using novel nanodevices featuring physics analogous to metaplasticity.

cs.NE↗

arXiv2021

Synaptic metaplasticity in binarized neural networks

Unlike the brain, artificial neural networks, including state-of-the-art deep neural networks for computer vision, are subject to "catastrophic forgetting": they rapidly forget the previous task when trained on a new one. Neuroscience suggests that biological synapses avoid this issue through the process of synaptic consolidation and metaplasticity: the plasticity itself changes upon repeated synaptic events. In this work, we show that this concept of metaplasticity can be transferred to a particular type of deep neural networks, binarized neural networks, to reduce catastrophic forgetting.

cs.NE↗

arXiv2021

Scaling Equilibrium Propagation to Deep ConvNets by Drastically Reducing its Gradient Estimator Bias

Equilibrium Propagation (EP) is a biologically-inspired counterpart of Backpropagation Through Time (BPTT) which, owing to its strong theoretical guarantees and the locality in space of its learning rule, fosters the design of energy-efficient hardware dedicated to learning. In practice, however, EP does not scale to visual tasks harder than MNIST. In this work, we show that a bias in the gradient estimate of EP, inherent in the use of finite nudging, is responsible for this phenomenon and that cancelling it allows training deep ConvNets by EP, including architectures with distinct forward and backward connections. These results highlight EP as a scalable approach to compute error gradients in deep neural networks, thereby motivating its hardware implementation.

cs.LG↗

arXiv2020

Implementation of Ternary Weights with Resistive RAM Using a Single Sense Operation per Synapse

The design of systems implementing low precision neural networks with emerging memories such as resistive random access memory (RRAM) is a significant lead for reducing the energy consumption of artificial intelligence. To achieve maximum energy efficiency in such systems, logic and memory should be integrated as tightly as possible. In this work, we focus on the case of ternary neural networks, where synaptic weights assume ternary values. We propose a two-transistor/two-resistor memory architecture employing a precharge sense amplifier, where the weight value can be extracted in a single sense operation. Based on experimental measurements on a hybrid 130 nm CMOS/RRAM chip featuring this sense amplifier, we show that this technique is particularly appropriate at low supply voltage, and that it is resilient to process, voltage, and temperature variations. We characterize the bit error rate in our scheme. We show based on neural network simulation on the CIFAR-10 image recognition task that the use of ternary neural networks significantly increases neural network performance, with regards to binary ones, which are often preferred for inference hardware. We finally evidence that the neural network is immune to the type of bit errors observed in our scheme, which can therefore be used without error correction.

cs.ET↗

arXiv2020

Scaling Equilibrium Propagation to Deep ConvNets by Drastically Reducing its Gradient Estimator Bias

Equilibrium Propagation (EP) is a biologically-inspired algorithm for convergent RNNs with a local learning rule that comes with strong theoretical guarantees. The parameter updates of the neural network during the credit assignment phase have been shown mathematically to approach the gradients provided by Backpropagation Through Time (BPTT) when the network is infinitesimally nudged toward its target. In practice, however, training a network with the gradient estimates provided by EP does not scale to visual tasks harder than MNIST. In this work, we show that a bias in the gradient estimate of EP, inherent in the use of finite nudging, is responsible for this phenomenon and that cancelling it allows training deep ConvNets by EP. We show that this bias can be greatly reduced by using symmetric nudging (a positive nudging and a negative one). We also generalize previous EP equations to the case of cross-entropy loss (by opposition to squared error). As a result of these advances, we are able to achieve a test error of 11.7% on CIFAR-10 by EP, which approaches the one achieved by BPTT and provides a major improvement with respect to the standard EP approach with same-sign nudging that gives 86% test error. We also apply these techniques to train an architecture with asymmetric forward and backward connections, yielding a 13.2% test error. These results highlight EP as a compelling biologically-plausible approach to compute error gradients in deep neural networks.

cs.NE↗

arXiv2020

Low Power In-Memory Implementation of Ternary Neural Networks with Resistive RAM-Based Synapse

The design of systems implementing low precision neural networks with emerging memories such as resistive random access memory (RRAM) is a major lead for reducing the energy consumption of artificial intelligence (AI). Multiple works have for example proposed in-memory architectures to implement low power binarized neural networks. These simple neural networks, where synaptic weights and neuronal activations assume binary values, can indeed approach state-of-the-art performance on vision tasks. In this work, we revisit one of these architectures where synapses are implemented in a differential fashion to reduce bit errors, and synaptic weights are read using precharge sense amplifiers. Based on experimental measurements on a hybrid 130 nm CMOS/RRAM chip and on circuit simulation, we show that the same memory array architecture can be used to implement ternary weights instead of binary weights, and that this technique is particularly appropriate if the sense amplifier is operated in near-threshold regime. We also show based on neural network simulation on the CIFAR-10 image recognition task that going from binary to ternary neural networks significantly increases neural network performance. These results highlight that AI circuits function may sometimes be revisited when operated in low power regimes.

cs.ET↗

arXiv2018

MicroMegascope

Atomic Force Microscopy (AFM) allows to reconstruct the topography of surface with a resolution in the nanometer range. The exceptional resolution attainable with the AFM makes this instrument a key tool in nanoscience and technology. The core of the set-up relies on the detection of the mechanical properties of a micro-oscillator when approached to a sample to image. Despite the fact that AFM is nowadays a very common instrument for research and development applications, thanks to the exceptional performances and the relative simplicity to use it, the fabrication of the micrometric scale mechanical oscillator is still a very complicated and expensive task requiring a dedicated platform. Being able to perform atomic force microscopy with a macroscopic oscillator would make the instrument more versatile and accessible for an even larger spectrum of applications and audiences. We present for the first time atomic force imaging with a centimetric oscillator. We show how it is possible to perform topographical images with nanometric resolution with a grams tuning fork. The images presented here are obtained with an aluminum tuning fork of centimeter size as sensor on which an accelerometer is glued on one prong to measure the oscillation of the resonator. In addition to the stunning sensitivity, by imaging both in air and in liquid, we show the high versatility of such oscillator. The set up proposed here can be extended to numerous experiments where the probe needs to be heavy and/or very complex as well as the environment.

physics.ins-det↗