arXiv subjects

Claudia Clopath

Publications and source records attributed to Claudia Clopath.

At least 19 recordsLinked to original sources

arXiv2026

Overcoming Rank Collapse in Feedback Alignment

Backpropagation (BP) is widely viewed as biologically implausible, in part because it requires feedback weights to be the transpose of forward weights for error propagation. Interestingly, when training a network with fixed random feedback weights to circumvent this issue, learning aligns the forward weights with the feedback weights, leading the backpropagated error signal to become an approximation of the standard gradient used by BP. This process, called Feedback Alignment (FA), occurs in MLPs and very shallow CNNs but does not scale well to deeper architectures. In this work, we first investigated differences between BP and FA models, trained on CIFAR10, specifically focusing on the effective rank of the signal. We found that the FA error has a considerably lower rank and hence is constrained to a lower-dimensional subspace compared to BP, limiting exploration of the parameter space. Motivated by this observation, we evaluated two mechanisms for increasing the effective dimensionality of FA: Muon, an optimiser that orthogonalises weight updates; and hidden activity normalisation, which promotes activation orthogonality. Across larger architectures and benchmarks, we find that these methods consistently improve over FA baselines, for example, on CIFAR100 with a Resnet-18, accuracy increases by 9 percentage points. Our results identify low-dimensional gradient dynamics as a key obstacle to scaling FA and suggest that inducing higher-dimensional update geometry is a promising route toward scaling alternatives to backpropagation.

cs.LG↗

arXiv2026

Cortico-cerebellar modularity as an architectural inductive bias for efficient temporal learning

The cerebellum and cerebral cortex form tightly coupled circuits thought to support flexible and efficient temporal processing. How this interaction shapes cortical learning dynamics, and whether such heterogeneous modularity can benefit artificial systems, remains unclear. Here, we augment a recurrent neural network (RNN) with a cerebellar-inspired feedforward module and evaluate the resulting architecture on temporal tasks of varying difficulty. The cortico-cerebellar RNN (CB-RNN) learns faster and reaches higher maximum performance than parameter-matched fully recurrent baselines across a variety of regimes. Crucially, freezing the recurrent core after minimal training and delegating subsequent learning to the cerebellar module preserves superior learning efficiency, suggesting the cerebellar module is a primary driver of efficiency and that the cortical network can largely function as a fixed reservoir. Our results suggest that heterogeneous modular architectures can act as a powerful structural inductive bias in neural systems.

q-bio.NC↗

arXiv2025

Uncertainty Prioritized Experience Replay

Prioritized experience replay, which improves sample efficiency by selecting relevant transitions to update parameter estimates, is a crucial component of contemporary value-based deep reinforcement learning models. Typically, transitions are prioritized based on their temporal difference error. However, this approach is prone to favoring noisy transitions, even when the value estimation closely approximates the target mean. This phenomenon resembles the noisy TV problem postulated in the exploration literature, in which exploration-guided agents get stuck by mistaking noise for novelty. To mitigate the disruptive effects of noise in value estimation, we propose using epistemic uncertainty estimation to guide the prioritization of transitions from the replay buffer. Epistemic uncertainty quantifies the uncertainty that can be reduced by learning, hence reducing transitions sampled from the buffer generated by unpredictable random processes. We first illustrate the benefits of epistemic uncertainty prioritized replay in two tabular toy models: a simple multi-arm bandit task, and a noisy gridworld. Subsequently, we evaluate our prioritization scheme on the Atari suite, outperforming quantile regression deep Q-learning benchmarks; thus forging a path for the use of uncertainty prioritized replay in reinforcement learning agents.

cs.LG↗

arXiv2025

Relational reasoning and inductive bias in transformers and large language models

Transformer-based models have demonstrated remarkable reasoning abilities, but the mechanisms underlying relational reasoning remain poorly understood. We investigate how transformers perform \textit{transitive inference}, a classic relational reasoning behavior from psychology which elicits inference about indirectly related items (e.g., if $A > B$ and $B > C$, then $A > C$). We compare in-weights learning (IWL) and in-context learning (ICL) behaviors and mechanisms on these tasks, and fine profoundly different patterns of generalization. IWL models learn a linear embedding, which leads to transitive inference as well as other behavioral effects present in humans and animals. ICL models, in contrast, are capable of learning to generalize transitively, but only do so when it is necessitated by the training data, otherwise learning a match-and-copy strategy. Interestingly, pre-training ICL models on in-context linear regression tasks that provide them with a latent linear representation is sufficient to make the ICL behaviors and internal representations qualitatively and quantitatively more like IWL. In order to test whether the same inference patterns are present across in large language models, we leverage a congruency paradigm which allows us to differentially probe IWL and ICL generalization patterns without access to their training data. We indeed see IWL reasoning leads to more transitive generalization than ICL. Moreover, we find that prompting the ICL models to use a linear mental map led to increased transitive inference over different geometric prompts. Together, these results reveal that both the training regime and the geometric structure of induced representations critically determine transformers capacity for transitive inference.

cs.LG↗

arXiv2025

Neural mechanisms of predictive processing: a collaborative community experiment through the OpenScope program

This review synthesizes advances in predictive processing within the sensory cortex. Predictive processing theorizes that the brain continuously predicts sensory inputs, refining neuronal responses by highlighting prediction errors. We identify key computational primitives, such as stimulus adaptation, dendritic computation, excitatory/inhibitory balance and hierarchical processing, as central to this framework. Our review highlights convergences, such as top-down inputs and inhibitory interneurons shaping mismatch signals, and divergences, including species-specific hierarchies and modality-dependent layer roles. To address these conflicts, we propose experiments in mice and primates using in-vivo two-photon imaging and electrophysiological recordings to test whether temporal, motor, and omission mismatch stimuli engage shared or distinct mechanisms. The resulting dataset, collected and shared via the OpenScope program, will enable model validation and community analysis, fostering iterative refinement and refutability to decode the neural circuits of predictive processing.

q-bio.NC↗

arXiv2024

Lifelong Reinforcement Learning via Neuromodulation

Navigating multiple tasks$\unicode{x2014}$for instance in succession as in continual or lifelong learning, or in distributions as in meta or multi-task learning$\unicode{x2014}$requires some notion of adaptation. Evolution over timescales of millennia has imbued humans and other animals with highly effective adaptive learning and decision-making strategies. Central to these functions are so-called neuromodulatory systems. In this work we introduce an abstract framework for integrating theories and evidence from neuroscience and the cognitive sciences into the design of adaptive artificial reinforcement learning algorithms. We give a concrete instance of this framework built on literature surrounding the neuromodulators Acetylcholine (ACh) and Noradrenaline (NA), and empirically validate the effectiveness of the resulting adaptive algorithm in a non-stationary multi-armed bandit problem. We conclude with a theory-based experiment proposal providing an avenue to link our framework back to efforts in experimental neuroscience.

cs.LG↗

arXiv2023

A study on the plasticity of neural networks

One aim shared by multiple settings, such as continual learning or transfer learning, is to leverage previously acquired knowledge to converge faster on the current task. Usually this is done through fine-tuning, where an implicit assumption is that the network maintains its plasticity, meaning that the performance it can reach on any given task is not affected negatively by previously seen tasks. It has been observed recently that a pretrained model on data from the same distribution as the one it is fine-tuned on might not reach the same generalisation as a freshly initialised one. We build and extend this observation, providing a hypothesis for the mechanics behind it. We discuss the implication of losing plasticity for continual learning which heavily relies on optimising pretrained models.

cs.LG↗

arXiv2023

When Does Re-initialization Work?

Re-initializing a neural network during training has been observed to improve generalization in recent works. Yet it is neither widely adopted in deep learning practice nor is it often used in state-of-the-art training protocols. This raises the question of when re-initialization works, and whether it should be used together with regularization techniques such as data augmentation, weight decay and learning rate schedules. In this work, we conduct an extensive empirical comparison of standard training with a selection of re-initialization methods to answer this question, training over 15,000 models on a variety of image classification benchmarks. We first establish that such methods are consistently beneficial for generalization in the absence of any other regularization. However, when deployed alongside other carefully tuned regularization techniques, re-initialization methods offer little to no added benefit for generalization, although optimal generalization performance becomes less sensitive to the choice of learning rate and weight decay hyperparameters. To investigate the impact of re-initialization methods on noisy data, we also consider learning under label noise. Surprisingly, in this case, re-initialization significantly improves upon standard training, even in the presence of other carefully tuned regularization techniques.

cs.LG↗

arXiv2023

Toward Next-Generation Artificial Intelligence: Catalyzing the NeuroAI Revolution

Neuroscience has long been an essential driver of progress in artificial intelligence (AI). We propose that to accelerate progress in AI, we must invest in fundamental research in NeuroAI. A core component of this is the embodied Turing test, which challenges AI animal models to interact with the sensorimotor world at skill levels akin to their living counterparts. The embodied Turing test shifts the focus from those capabilities like game playing and language that are especially well-developed or uniquely human to those capabilities, inherited from over 500 million years of evolution, that are shared with all animals. Building models that can pass the embodied Turing test will provide a roadmap for the next generation of AI.

cs.AI↗

arXiv2022

Maslow's Hammer for Catastrophic Forgetting: Node Re-Use vs Node Activation

Continual learning - learning new tasks in sequence while maintaining performance on old tasks - remains particularly challenging for artificial neural networks. Surprisingly, the amount of forgetting does not increase with the dissimilarity between the learned tasks, but appears to be worst in an intermediate similarity regime. In this paper we theoretically analyse both a synthetic teacher-student framework and a real data setup to provide an explanation of this phenomenon that we name Maslow's hammer hypothesis. Our analysis reveals the presence of a trade-off between node activation and node re-use that results in worst forgetting in the intermediate regime. Using this understanding we reinterpret popular algorithmic interventions for catastrophic interference in terms of this trade-off, and identify the regimes in which they are most effective.

stat.ML↗

arXiv2022

Current State and Future Directions for Learning in Biological Recurrent Neural Networks: A Perspective Piece

We provide a brief review of the common assumptions about biological learning with findings from experimental neuroscience and contrast them with the efficiency of gradient-based learning in recurrent neural networks. The key issues discussed in this review include: synaptic plasticity, neural circuits, theory-experiment divide, and objective functions. We conclude with recommendations for both theoretical and experimental neuroscientists when designing new studies that could help bring clarity to these issues.

q-bio.NC↗

arXiv2021

Spectral Normalisation for Deep Reinforcement Learning: an Optimisation Perspective

Most of the recent deep reinforcement learning advances take an RL-centric perspective and focus on refinements of the training objective. We diverge from this view and show we can recover the performance of these developments not by changing the objective, but by regularising the value-function estimator. Constraining the Lipschitz constant of a single layer using spectral normalisation is sufficient to elevate the performance of a Categorical-DQN agent to that of a more elaborated \rainbow{} agent on the challenging Atari domain. We conduct ablation studies to disentangle the various effects normalisation has on the learning dynamics and show that is sufficient to modulate the parameter updates to recover most of the performance of spectral normalisation. These findings hint towards the need to also focus on the neural component and its learning dynamics to tackle the peculiarities of Deep Reinforcement Learning.

cs.LG↗

arXiv2020

Continual Reinforcement Learning with Multi-Timescale Replay

In this paper, we propose a multi-timescale replay (MTR) buffer for improving continual learning in RL agents faced with environments that are changing continuously over time at timescales that are unknown to the agent. The basic MTR buffer comprises a cascade of sub-buffers that accumulate experiences at different timescales, enabling the agent to improve the trade-off between adaptation to new data and retention of old knowledge. We also combine the MTR framework with invariant risk minimization, with the idea of encouraging the agent to learn a policy that is robust across the various environments it encounters over time. The MTR methods are evaluated in three different continual learning settings on two continuous control tasks and, in many cases, show improvement over the baselines.

cs.LG↗

arXiv2019

Learning spatiotemporal signals using a recurrent spiking network that discretizes time

Learning to produce spatiotemporal sequences is a common task that the brain has to solve. The same neural substrate may be used by the brain to produce different sequential behaviours. The way the brain learns and encodes such tasks remains unknown as current computational models do not typically use realistic biologically-plausible learning. Here, we propose a model where a spiking recurrent network of excitatory and inhibitory biophysical neurons drives a read-out layer: the dynamics of the driver recurrent network is trained to encode time which is then mapped through the read-out neurons to encode another dimension, such as space or a phase. Different spatiotemporal patterns can be learned and encoded through the synaptic weights to the read-out neurons that follow common Hebbian learning rules. We demonstrate that the model is able to learn spatiotemporal dynamics on time scales that are behaviourally relevant and we show that the learned sequences are robustly replayed during a regime of spontaneous activity.

q-bio.NC↗

arXiv2019

Cortical credit assignment by Hebbian, neuromodulatory and inhibitory plasticity

The cortex learns to make associations between stimuli and spiking activity which supports behaviour. It does this by adjusting synaptic weights. The complexity of these transformations implies that synapses have to change without access to the full error information, a problem typically referred to as "credit-assignment". However, it remains unknown how the cortex solves this problem. We propose that a combination of plasticity rules, 1) Hebbian, 2) acetylcholine-dependent and 3) noradrenaline-dependent excitatory plasticity, together with 4) inhibitory plasticity restoring E/I balance, effectively solves the credit assignment problem. We derive conditions under-which a neuron model can learn a number of associations approaching its theoretical capacity. We confirm our predictions regarding acetylcholine-dependent and inhibitory plasticity by reanalysing experimental data. Our work suggests that detailed cortical E/I balance reduces the dimensionality of the problem of associating inputs with outputs, thereby allowing imperfect "supervision" by neuromodulatory systems to guide learning effectively.

q-bio.NC↗

arXiv2019

Policy Consolidation for Continual Reinforcement Learning

We propose a method for tackling catastrophic forgetting in deep reinforcement learning that is \textit{agnostic} to the timescale of changes in the distribution of experiences, does not require knowledge of task boundaries, and can adapt in \textit{continuously} changing environments. In our \textit{policy consolidation} model, the policy network interacts with a cascade of hidden networks that simultaneously remember the agent's policy at a range of timescales and regularise the current policy by its own history, thereby improving its ability to learn without forgetting. We find that the model improves continual learning relative to baselines on a number of continuous control tasks in single-task, alternating two-task, and multi-agent competitive self-play settings.

cs.LG↗

arXiv2018

A High GOPs/Slice Time Series Classifier for Portable and Embedded Biomedical Applications

Nowadays a diverse range of physiological data can be captured continuously for various applications in particular wellbeing and healthcare. Such data require efficient methods for classification and analysis. Deep learning algorithms have shown remarkable potential regarding such analyses, however, the use of these algorithms on low-power wearable devices is challenged by resource constraints such as area and power consumption. Most of the available on-chip deep learning processors contain complex and dense hardware architectures in order to achieve the highest possible throughput. Such a trend in hardware design may not be efficient in applications where on-node computation is required and the focus is more on the area and power efficiency as in the case of portable and embedded biomedical devices. This paper presents an efficient time-series classifier capable of automatically detecting effective features and classifying the input signals in real-time. In the proposed classifier, throughput is traded off with hardware complexity and cost using resource sharing techniques. A Convolutional Neural Network (CNN) is employed to extract input features and then a Long-Short-Term-Memory (LSTM) architecture with ternary weight precision classifies the input signals according to the extracted features. Hardware implementation on a Xilinx FPGA confirm that the proposed hardware can accurately classify multiple complex biomedical time series data with low area and power consumption and outperform all previously presented state-of-the-art records. Most notably, our classifier reaches 1.3$\times$ higher GOPs/Slice than similar state of the art FPGA-based accelerators.

cs.LG↗

arXiv2018

Continual Reinforcement Learning with Complex Synapses

Unlike humans, who are capable of continual learning over their lifetimes, artificial neural networks have long been known to suffer from a phenomenon known as catastrophic forgetting, whereby new learning can lead to abrupt erasure of previously acquired knowledge. Whereas in a neural network the parameters are typically modelled as scalar values, an individual synapse in the brain comprises a complex network of interacting biochemical components that evolve at different timescales. In this paper, we show that by equipping tabular and deep reinforcement learning agents with a synaptic model that incorporates this biological complexity (Benna & Fusi, 2016), catastrophic forgetting can be mitigated at multiple timescales. In particular, we find that as well as enabling continual learning across sequential training of two simple tasks, it can also be used to overcome within-task forgetting by reducing the need for an experience replay database.

cs.AI↗