NeuroTorch: A Python library for neuroscience-oriented machine learning
Bibliographic record
Abstract
Machine learning (ML) has become a powerful tool for data analysis, leading to significant advances in neuroscience research. While ML algorithms are proficient in general-purpose tasks, their highly technical nature often hinders their compatibility with the observed biological principles and constraints in the brain, thereby limiting their suitability for neuroscience applications. In this work, we introduce NeuroTorch, a comprehensive ML pipeline specifically designed to assist neuroscientists in leveraging ML techniques using biologically inspired neural network models. NeuroTorch enables the training of recurrent neural networks equipped with either spiking or firing-rate dynamics, incorporating additional biological constraints such as Dale's law and synaptic excitatory-inhibitory balance. The pipeline offers various learning methods, including backpropagation through time and eligibility trace forward propagation, aiming to allow neuroscientists to effectively employ ML approaches. To evaluate the performance of NeuroTorch, we conducted experiments on well-established public datasets for classification tasks, namely MNIST, Fashion-MNIST, and Heidelberg. Notably, NeuroTorch achieved accuracies that replicated the results obtained using the Norse and SpyTorch packages. Additionally, we tested NeuroTorch on real neuronal activity data obtained through volumetric calcium imaging in larval zebrafish. On training sets representing 9.3 minutes of activity under darkflash stimuli from 522 neurons, the mean proportion of variance explained for the spiking and firing-rate neural network models, subject to Dale's law, exceeded 0.97 and 0.96, respectively. Our analysis of networks trained on these datasets indicates that both Dale's law and spiking dynamics have a beneficial impact on the resilience of network models when subjected to connection ablations. NeuroTorch provides an accessible and well-performing tool for neuroscientists, granting them access to state-of-the-art ML models used in the field without requiring in-depth expertise in computer science.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.006 |
| Meta-epidemiology (narrow) | 0.003 | 0.002 |
| Meta-epidemiology (broad) | 0.002 | 0.003 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.005 | 0.004 |
| Research integrity | 0.001 | 0.005 |
| Insufficient payload (model declined to judge) | 0.053 | 0.027 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".