One of the worst airport experiences in my life not withstanding, I just had one of the best travel experiences of my life!

It started with ICML in Seoul, where I presented my work on unitary convolutions with Luisa and John. I've presented the poster a couple times already, but I was still wipped at the end of this one. But the academics are the boring part anyway, let's talk about THE CITY and THE PEOPLE.

Starting with the city, it makes a strong first impression. Incheon airport is gorgeous, and with the help of some locals I was able to pretty seemlessly get my transit card and ride a subway that puts the US to shame. I stayed in Gangnam-Gu, which is on the South side of the Han river. I stayed like a block away from some of the famous tombs. Apparently Gangnam-Gu is the young / new area. I was also in walking distance of the Coex convention center, and man that place itself is also sick. Its on top of a tunnel system connecting you to a shopping mall, acquarium, big library, and transit to all of the nice tourist sights. Of those tourist attractions, Luisa and I saw the palaces, the N Seoul Tower, some old villages with some vintage architecture, and the myeongdong night market. I didn't get to see SNU... maybe next time! Regardless. How magical!

As always. the people is what makes the conference special. This time around, I was less focused on getting into graduate school and more focused on just learning new things from new people. I learned a lot about PLMs, diffusion language models and categorical flows maps, and robotics. I felt a lot of pride seeing timely work from people in my department at Harvard CS! It was great meeting you Kiho, Moritz, Rosie and Jay! And to Kiho and Moritz, I may see you in Melanie's lab soon! :D. Beyond Harvard, I enjoyed getting coffee with John and meeting Jeffrey and learning more about the AI 4 drug discovery industry, learning about categorical flow maps from Floor, and nerding out about graphs with Lucky.

Following Seoul, I went to London for LOGML. There were great talks from Soledad, Stefanie, and Nicholas Sharp. Nich was especially exciting because my paper on unitary convolutions heavily depended on the robust laplacian (which he invented), and I got to talk to him about it! He announced during his talk that he got a faculty position at U Dub, and I think it is super well deserved!

Research wise, I've been having a lot of fun reading about and implementing diffusion forcing for my project with Robin and Tommy. My mode volume project with Robin and Bo got a spotlight at the TagDS workshop, and I think Melanie and I are going to target ICLR with the splats!!! So all in all, things are going well even if things were a bit slow when I was travelling.

Outside of research, I managed to do some fun stuff before I left and during my flights. I finished heated rivarly, ponyo, and the boy and the heron. Brady Tkachuk was also traded from the Ottawa Senators to the Florida Panthers. As I get older I have less judgement for professional athletes becoming "traitors" to the team that drafted them and doing whats best for them and their families. It was never really justified honestly, I was just caught up in the fandom of it all. I would not like it very much if one day Melanie said she was trading me to UC Berkeley for 2 rolls of tape and a bag of pucks. As Jake reminded me this week, I am living the dream going to school in Boston :).

All of that being said... One day there will be a book about the soft tampering that happened during the 2026 Olympic games amongst the team USA players. Bonus point to Mitch Marner soft tampering with Vegas during the previous years four nations faceoff.

Oh and I started reading my hero academia.