Linkage

You’re probably familiar with machine-learning-based translation between natural languages, based on finding patterns in large datasets of known translations, for instance as used by Google translate. Now the Xena people are trying to use the same methods to convert LaTeX-formatted natural-language descriptions of mathematical propositions into the formal language used by the Lean theorem prover (\(\mathbb{M}\)).

Pretty shadertoy flythrough of the Laves graph (\(\mathbb{M}\), via).

The Computational Complexity blog takes on the question of the conference-based publishing culture in computer science (\(\mathbb{M}\)), and whether in-person vs virtual vs hybrid conferences can really be said to be working, now that we have enough experience going back and forth between these modalities and the novelty of the virtual and hybrid formats has worn off.

Many minimal triangulations of “manifolds like the octonionic projective plane” (\(\mathbb{M}\)). Blog post by Gil Kalai based on a new preprint by Alexander Gaifullin. Gaifullin conjectures that these all are the octonionic projective plane (not merely “like it”). Gil’s post connects this to several other extremal problems, mostly in polyhedral combinatorics but also including the existence of a girth-5 degree-57 Moore graph.

The Journal of Vascular and Interventional Radiology retracts its crossword puzzle (\(\mathbb{M}\), via, via2). No explanation why yet, but I suspect it’s not because of falsified experimental data.

Today’s simple and somewhat obvious geometry observation (\(\mathbb{M}\)), made while trying to fit old CD cover art into sleeves of a CD binder for more compact storage: if you’re trying to fit a flexible sheet into a flexible sleeve that is only just barely big enough to hold it, it doesn’t help to bend both into a convex curve; they will still be too tight. Instead, if you bend the sheet into a compound curve, while letting the sleeve fall into a convex shape surrounding it, you can make more room.

Olligobber implements combinatory logic in \(\TeX\) macros (\(\mathbb{M}\)).

The graph minor theorem meets algebra: in the Notices, Eric Ramos explains a conjectured category-theoretic generalization of the fact that graph minors form a well-quasi-ordering. Via a long multi-post thread by John Baez.

New draft book on lower bounds in complexity theory (\(\mathbb{M}\), via), by Demaine, Gasarch, and Hajiaghayi, intended as a replacement for Garey and Johnson’s 1979 NP-completeness book.

Why A4? – the mathematical beauty of paper size (\(\mathbb{M}\)). You probably already knew about the self-reproducing shape of \(1\times\sqrt2\) rectangles when folded in half, but the link is a nice explainer of why A4 paper has such odd-looking dimensions for those who didn’t know.

Talk slides for my papers at the just-concluded Canadian Conference on Computational Geometry (\(\mathbb{M}\)):

• Orthogonal dissection into few rectangles (see earlier post)

• Locked and unlocked smooth embeddings of surfaces (see earlier post)

• Reflections in an octagonal mirror maze (see earlier post)

An update on the “coffee near Ryerson” map from my recent post (\(\mathbb{M}\)): I tried four cafés, only one of which was highlighted by a line on the map: Black Bear, Page One, Hailed, and Mast. I would have also tried Le Génie but it was closed mornings. The best coffee was at Hailed, but it’s tiny, with only outdoor seating. Second-best coffee, and the best space to hang out to drink the coffee, was Mast. The others did not disappoint but were not as good as Hailed and Mast.

The toxic culture of rejection in computer science (\(\mathbb{M}\), via), Edward Lee in the ACM SIGBED blog. I disagree with the post’s preference for incrementalism over novelty, but I agree that there’s big price for being too selective. Beyond frustrating everyone, I think it leads to dominance of trendiness and in-groups over significance, progress, originality, and depth. And though that may be good for those in the trendy in-groups, it’s not good for the field.

Accusations of research misconduct against Princeton gerrymandering researcher Sam Wang, made by a Republican political operative, have been found to be “without merit” by the university (\(\mathbb{M}\), via).

How origami is engineering new technological opportunities (\(\mathbb{M}\)). Interview with mechanical engineer Sachiko Ishida of Meiji University on applications of folded structures in engineering, and where origami engineering is headed.