Neural Reckoning Group

Latest from Bluesky

Dan Goodman

Unifying principles of intelligent systems.

Yang Chu

Robust and flexible perception under distribution shift

Hardik Rajpal

Self-organization and generalisation in learning.

Greta Horvathova

Information bottlenecks in intelligent systems.

Jatin Sharma

Interpretable machine learning and neuroscience.

AbdelQader AlKilany

Neuromodulation and neuromorphic computing.

Nicolás Dundov Muñoz

Neural architectures, information theory and neuromorphic computing.

Iñigo Lara

Physical neuromorphic computing.

Kyle McMillan

Local learning and neuromorphic computing.

Anna Obure

Developmental learning in multimodal intelligence.

Yishu Zhang

Mathematical and computational techniques to understand the theoretical foundation of neural heterogeneity.

Marcus Ghosh

Discovering how brain structure shapes function.

The Neural Reckoning Group is led by Dan Goodman at Imperial College London.

We aim to find unifying principles underlying intelligent systems, including biological systems such as the brain and artificial systems. We use theoretical and computational approaches.

We are particularly interested in spiking neural networks and the role they play in sensory processing. Machine learning is essential to our work, as we believe that only by understanding how the brain copes with messy, real-world complexity can we hope to understand what makes it unique. A key part of our work is neuroinformatics, building open source software packages to make our methods freely available to all. For a brief overview of our interests, see the selection of papers, software and organisations below.

A good place to start to get a feel for our current topics is Dan Goodman's Brain Inspired podcast interview.

We maintain a short list of resources for learning computational neuroscience that might be useful to students and people new to the field.

We are also interested in reforming and improving the way we do science.

Recent work

Paper last updated 2026-08-12:

Habashy KG, Evans BD, Goodman DFM, Bowers JS
Factorization and spatial encodings: a hypothesis about the foundations of the genomic code.
Preprint
Paper last updated 2026-08-07:

Ghosh M, Goodman DFM (2026)
Partial recurrence enables robust and efficient computation.
Communications AI & Computing
Talk last updated 2026-07-15:

Learning with spikes
Dan Goodman at FENS-Chen Institute summer school lecture (Cambridge) (2026)
Paper last updated 2026-07-14:

Rajpal H, Goodman DFM
Emergent Generalization by Representation Learning in Artificial Neural Networks.
Preprint
Paper last updated 2026-07-13:

AlKilany A, Goodman DFM
Neuromodulation enhances the capability and efficiency of spiking neural networks.
Preprint

Selected publications

Sun P, Su Z, Achterberg J, Indiveri G, Goodman DFM, Akarca D (2026)
Algorithm-hardware co-design of neuromorphic networks with dual memory pathways.
Nature Machine Intelligence
Béna G, Goodman DFM (2025)
Dynamics of specialization in neural modules under resource constraints.
Nature Communications
Ghosh M, Béna G, Bormuth V, Goodman DFM (2024)
Nonlinear fusion is optimal for a wide class of multisensory tasks.
PLoS Computational Biology
Perez-Nieves N, Leung VCH, Dragotti PL, Goodman DFM (2021)
Neural heterogeneity promotes robust learning.
Nature Communications
Zenke F, et al. (2021)
Visualizing a joint future of neuroscience and neuromorphic engineering.
Neuron
Achakulvisut T, et al. (2021)
Towards democratizing and automating online conferences: lessons from the Neuromatch conferences.
Trends in Cognitive Sciences
Stimberg M, Brette R, Goodman DFM (2019)
Brian 2, an intuitive and efficient neural simulator.
eLife

Other things we do

We build open source software like Brian, and organisations like Neuromatch and SNUFA.

A Python simulator for spiking neural networks.

Inclusive networking and education for neuroscience.

Organisation for neuroscientists interested in spiking neural networks.

Latest video

See more videos here.