iiinova — where your digital self keeps evolving

It's that time of year where the view from my window gets quite distracting. #photography

Rich, golden leaves in the light of the setting sun.

Dan Goodman @neural-reckoning.org · 01/10/2026

This was my first time going to #BernsteinConference and I really enjoyed it. Would definitely recommend going next year. Particularly enjoyed the workshops and posters. Overall vibe was 👇

Gigantic wall mural of a jellybean playing saxophone.

311314

Dan Goodman @neural-reckoning.org · 01/10/2026

I hardly ever take flights any more, but I have to admit that coming in to land in London at sunset with a window seat is a pretty spectacular experience.

1170

Dan Goodman @neural-reckoning.org · 30/09/2026

My contribution to the black and white photography in fog (well in this case, from inside a cloud).

110

Dan Goodman @neural-reckoning.org · 25/09/2026

Nicolas Dundov has poster 4-38 (first floor) on Wednesday (14:00-15:30) on "Blood and modularity: a brain-inspired approach to functional disentangled neural representations" neural-reckoning.org/nicolas_dund...

030

Dan Goodman @neural-reckoning.org · 25/09/2026

Abdal AlKilany has poster 4-3 (ground floor) on Wednesday (14:00-15:30) on "Neuromodulation improves robust and efficient sensory processing in spiking neural networks" neural-reckoning.org/abdelqader_a... Preprint: neural-reckoning.org/pub_neuromod... Talk: neural-reckoning.org/talk_2026_ne...

130

Dan Goodman @neural-reckoning.org · 25/09/2026

Yishu Zhang has poster P1-44 (first floor) on Tuesday (16:30-18:00) on "Do heterogeneous brains remember better?" neural-reckoning.org/yishu_zhang....

120

Dan Goodman @neural-reckoning.org · 24/09/2026

Not a terrible evening to go for a run home though the city.

Golden purple sunset over one of the lakes in Hyde park.

Golden purple sunset over the Royal Albert Hall

040

Dan Goodman @neural-reckoning.org · 07/09/2026

Dappled sunlight in the woods is challenging to photograph. Most of my shots didn't come out that well, but I kinda like this one. #photography

A single leaf glowing luminous green with sunlight from behind. The background is very dark with a few indistinct shapes. You can see the veins on the leaf.

0160

Dan Goodman @neural-reckoning.org · 03/09/2026

#SpikingNeuralNetwork people - this year's SNUFA workshop is live! Talks by: ⭐ Eugene Izhikevich ⭐ @giuliadangelo.bsky.social ⭐ Mihai Petrovici ⭐ Susanne Schreiber Submit your abstracts by Sept 25th. Registration is open! More info at: snufa.net/2026/ cc @fzenke.bsky.social @albada.bsky.social

02416

Dan Goodman @neural-reckoning.org · 29/08/2026

This was totally worth staying up until 2am for. Word frequency analysis for the abstracts of my papers over time. You can read off my career from that, including my gradual reduction of interest in SNNs and then their sudden resurgence when surrogate gradient descent appeared.

Word frequency analysis over time of words that appear in the abstracts of my papers. Too much data to describe it all, but some of the trends are that model was strong up until about 2019 and has started to wane a bit. Neural and network have stayed pretty strong throughout, but particularly since 2019. Neuron was strong throughout. Words relating to the auditory system were up and down until about 2023 and have now vanished almost entirely. Task came in as a strong winner in the last few years. Method had a brief explosion around 2021-2023. Data was middling from 2013-2022 and now largely gone. Apparently I don't care about data any more.

0190

Dan Goodman @neural-reckoning.org · 23/08/2026

There's so many beautiful parks in London I sometimes forget about parkland walk, but I shouldn't because it's absolutely lovely. #photography #london

A couple holding hands and walking through dappled sunlight under a bridge in parkland walk. The sand from the path has been disturbed by runners and so there are streaks of light shining through the dust against the dark background of the tunnel under the bridge.

An artist spray painting against a wall in parkland walk. He's wearing a baseball cap, shades and a colourful mask to protect against the (admittedly pretty stinky) fumes.

Tall trees along the edge of the parkland walk path, dappled sunlight here and there.

View from under one of the bridge tunnels in the parkland walk. Colourful graffiti on the walls, lit up by a shaft of sunlight from the low sun from outside. Outside the tunnel you can see paths and trees.

0100

Dan Goodman @neural-reckoning.org · 18/08/2026

Paddington station. Which is better, night or day? #photography

Paddington station roof from inside during the day. Black and white, high contrast, striations.

Paddington at night. High contrast, dark ceiling, many lights of different colours.

370

Dan Goodman @neural-reckoning.org · 12/08/2026

My setup was not quite so professional!

Dan on a balcony with the sun behind him about to set behind a building. Visible is a camera (with two ND filters stacked), a bottle of suncream, a coke zero and a pair of sunglasses.

010

Dan Goodman @neural-reckoning.org · 12/08/2026

The results. I didn't have the right filter for my bigger zoom lens so quite low resolution, but still pretty happy with that.

Sequence of 12 images of the sun in various stages of eclipse.

2100

Dan Goodman @neural-reckoning.org · 12/08/2026

The glasses I ordered arrived just in time, and they actually delivered twice as many as I'd ordered so I got to spread the joy to a few lucky neighbours. #eclipse

1170

Dan Goodman @neural-reckoning.org · 11/08/2026

Any guesses as to what I'm doing at 7.15 pm BST taking a photo like this? 😂

Photo of the sun between some trees, and a very brightly outlined side of my head

170

Dan Goodman @neural-reckoning.org · 03/08/2026

Sometimes a lazy river can feel a bit like an alien landscape. #photography

A cluster of branches and roots fallen into a river.

The bank of a river with all the roots of a tree visible, and reflected in the river. It's a long exposure shot and little objects floating in the water shine bright as they pass through dappled sunlight creating streaks of light in the water.

A branch in a dark river with a few spots of dappled sunlight. In one of the spots of dappled sunlight a little group of water boatmen are clustered together, presumably enjoying the sun.

170

Dan Goodman @neural-reckoning.org · 16/07/2026

Modulation (top curve, red) also lets us reduce firing rates by orders of magnitude without hurting performance, unlike unmodulated networks (bottom curve, blue). Fewer spikes means less energy. Important both for real brains and neuromorphic devices!

130

Dan Goodman @neural-reckoning.org · 16/07/2026

In a neuromorphic or machine learning context, neuromodulation is very parameter efficient. This figure shows every variant of the modulated and unmodulated models in our paper, comparing parameter count versus accuracy. You can see that the frontier of modulated networks is much higher.

130

Dan Goodman @neural-reckoning.org · 16/07/2026

Our model allows for setting a number of neuromodulator types, that can have interactions between each other, and making neuromodulator release diffuse in space and time. This actually turns out to further improve performance!

120

Dan Goodman @neural-reckoning.org · 16/07/2026

And we can understand how it's doing it: it dials up the sensitivity when the noise level is low, and dials it down when the noise is high. This sort of dynamic gain control is the "listening in the dips" strategy that has been hypothesised to be used by humans. We didn't put this in, it learned it!

120

Dan Goodman @neural-reckoning.org · 16/07/2026

The performance enhancement is particularly large in a noisy background, in the range where human speech recognition is much better than state-of-the-art automatic speech recognition systems.

120

Dan Goodman @neural-reckoning.org · 16/07/2026

With this model, performance at a variety of tasks goes up (in the picture, an auditory task). A small network with modulation performs much better than a much larger network without modulation. We can also see here that modulation at medium timescales is best. Spatial scale not very important here.

120

Dan Goodman @neural-reckoning.org · 16/07/2026

Our model is simple at the core: we let a modulatory neuron modify the underlying parameters of another neuron (like threshold or time constant), and train end to end. We can make this more realistic in various ways (different types of neuromodulators, spatial release and diffusion) - see later.

120

Dan Goodman @neural-reckoning.org · 18/06/2026

The hardware version gets higher throughput (4x better) and energy efficiency (5x better) at comparable area.

120

Dan Goodman @neural-reckoning.org · 18/06/2026

We see that across a range of tasks we get state-of-the-art performance at a lower memory cost (M) compared to other models (look for the stars in the figure).

110

Dan Goodman @neural-reckoning.org · 18/06/2026

We think of this as a biologically inspired fast-slow structure where the SNNs are acting fast, and the memory buffer (a Legendre memory unit) acts as a slow working memory. We co-designed this to work as a neural model and in neuromorphic hardware.

110

Dan Goodman @neural-reckoning.org · 12/04/2026

Second joyful nature photo for today to celebrate election result in Hungary. #photography

Tree in bloom. White flowers glow luminously against a dark grey cloud background.

070

Dan Goodman @neural-reckoning.org · 10/04/2026

The kids had fun making paths on the ground in the local wood. #photography

Woodland with rich brown floor covered in leaves and sticks, and sun shining bright green through foliage above. On the floor there are trails of smoothed out mud lined with sticks forming paths.

642532

Dan Goodman @neural-reckoning.org · 08/04/2026

Not normally into bird #photography but quite pleased that I managed to get a shot of a woodpecker! Even managed to get a video of it pecking.

Woodpecker on the side of a tree with blue sky behind and some out of focus green leaves in front.

0142

Dan Goodman @neural-reckoning.org · 13/03/2026

In this sort of case I often think of Kent Anderson's hilarious article on "102 things journal publishers do" (to add value). One of which is to have a fancy office. That's not needed for science.