Calendar

June 2026

July 2026

MonTueWedThuFriSatSun12345678910111213141516171819202122232425262728293031

Nagoya, Aichi

Before the Typhoon

An evening talk with Professor Kato (an aquarium tip and an incoming typhoon), getting soaked to the bone leaving the lab, a late-night sushi dinner — and a map of the cyclone heading for Nagoya.

Weather: 21 °C, raining and thundering · Lunch: ebikatsu (delicious) · Dinner: sushi · Sleep: good, I just could have used more · Music: Neverhood OST (for some reason this is what I felt like listening to)

I need to do something about my sleep. Lately I’ve fallen so deep into the work that I’ve forgotten to explore the city and the country. Then again, I am here for research, so I shouldn’t mind it that much.

Right now my biggest trouble is collecting the rent and utilities — alongside the differential geometry, of course. But if the part I’m hunkered down here for works out, then again some huge reduction might come out of it.

Meanwhile the code kept running, of course, with the benchmarks and the measurements. I figured out a few things about what could be simplified and how to reduce the parameter count. Once again, astonishing results came; I hope my calculations pan out.

An evening talk with Kato

In the evening I stayed late, until about half past eight. Kato came in to talk with Sakuma about the progress of their own project, but he exchanged a few words with me too. For one, he liked the joint work with Zoli; I sent the papers, and I’ll send further information about our progress so far.

I showed him my results. Obviously these are initial results, but he liked what’s there so far, and said it could be taken further by combining the two kinds of neural network — and toward generating controllers based on hypergraph structures. Genius ideas.

I also showed him the COS conference organised by Széchenyi István University, and which journals the papers submitted there can be taken further into. So, on one hand, we’re already thinking about joint papers; on the other, he’s considering having his own students submit there. Really, what would rule it out? Although it’s an in-house conference, anyone can submit — last time the Chinese and just anyone from around the world submitted too. I also reported that I’d shown his student’s topic to one of my colleagues, in case they find some common ground.

So we had quite a good chat about this. After all, this is how you build connections. After that I asked what he could recommend for the weekend — apart from my research, of course. He said there’s an aquarium in Nagoya worth seeing. I jumped at the idea: I love aquariums.

But he told me it’s not worth going tomorrow, because the typhoon is coming, and it’ll flood the port. When he presented this as a fact, the way we at home say “the sun will shine and there’ll be a small drought,” I was reassured. I learned that although Nagoya is in a bay, the current between the Sea of Japan and the ocean runs right here. So typhoons are frequent — they happen five times a year even. Now it’s supposedly going to flood the port too, so it’s not worth heading that way right now. He also said Nagoya isn’t an earthquake-prone place, though the next big one (after Kobe and Fukushima) is supposedly expected right here.

Before the cyclone

Then, as I stepped out of the lab, I experienced what Kato was talking about. “Raining cats and dogs” is an understatement — it gained a whole new meaning here. On top of that, since it was late, I was hurrying to catch a place to eat. Just in the time it took to get back to the dorm, I got soaked to the bone — even though I had an umbrella: around me the water poured in streams (a river). It’s a wonder the drainage system works so well around here.

AI comic: the downpour before the typhoon
AI comic: the downpour before the typhoon

Really, by the time I got back to my room (I still live on the 5th floor, no elevator — that’s how I managed to get up in two minutes), there was only time to start the next benchmark and go eat. Luckily the sushi place — where I ate last week too — was still taking orders until 21:30, so I got to eat. I waited a bit, then got my table — this time I sat at a chair, not in a box. So I got my sushi straight to hand. I wolfed everything down again, but what did I eat?

  • onigiri from fresh tuna slices;
  • sushi with squid, a little wasabi;
  • seasonal teriyaki fish with rice (it’s become my personal favourite);
  • Asahi beer.

There’s nothing like the sushi they make here from the fresh seaweed leaf: crispy, as if it was just picked. I know it’s hard to do this at home — but that’s obviously why eating their own Japanese food in Japan is a different experience. Once I’d eaten, I went home, and this time I didn’t get as soaked: maybe because I didn’t rush, so there was no opportunity to.

As I looked at the map, a fairly large cyclone really is heading this way. I’m already getting notifications from everywhere to prepare. I’ll add: it’s cold and damp, you simply can’t set the AC right. And this is in huge contrast with the situation at home, where it’s 39 °C.

The approaching typhoon on the map
The approaching typhoon on the map

Photos

2026-06-26 01:31
2026-06-26 01:31
2026-06-26 01:34
2026-06-26 01:34
2026-06-26 01:34
2026-06-26 01:35
2026-06-26 01:35
2026-06-26 01:35
2026-06-26 01:37
2026-06-26 01:37
2026-06-26 01:37
2026-06-26 01:38
2026-06-26 01:38
2026-06-26 01:38
2026-06-26 01:38
2026-06-26 01:38
2026-06-26 01:40
2026-06-26 01:41
2026-06-26 01:41
2026-06-26 01:41
2026-06-26 16:58
2026-06-26 19:16
2026-06-26 19:23
2026-06-26 21:16
2026-06-26 21:21
2026-06-26 21:23
2026-06-26 21:48
2026-06-26 21:48
2026-06-26 21:48
2026-06-26 21:49
2026-06-26 21:49
Contents