Why is AI slop so frustrating to read?
AI writing often obscures main ideas and invents awkward phrases, making sentences more complex than necessary.
Many of these slop patterns violate writing principles established in English writing books.
In our benchmark, AI writing breaks these principles far more often than human writing, with Claude Code replies by far the worst offenders.
We show examples of the most common slop patterns in this thread
Surprising: Any orthogonal matrix Q
(QᵀQ = QQᵀ = I)
can produce a Doubly-Stochastic matrix S
(S𝟏 = 𝟏 & 𝟏ᵀS = 𝟏ᵀ)
in one shot. Just let Sₘₙ = | Qₘₙ |²
This class of doubly-stoch matrices maps vectors x to convex combination of permutations of x's elements.
Holy cr*p! @kirancodes explains how to get formal program semantics that are provably equivalent to the executed assembly🤯🤯🤯
That’s a real dreamer in action.
Just wait 10 years, every CS student will study her refinement theorem
Only at #LeanNYC
If you look in a mirror and say “isomonodromic deformations” five times fast, a differential geometer with a very thick mustache appears in your dreams that night and explains how the universe works to you.
I recently set out to 100x myself, and I think I’ve finally reached 10x consistency! Not 100x yet, but not bad!
I genuinely don’t think this would have been possible with the wisdom of engineers like @poteto and @copyconstruct. Thank you, sincerely. Coding is fun again!
How to effectively debug with agents when you no longer read or understand the code
In this post, I propose a debugging framework for agents that doesn’t rely on human knowledge of code to get to the right “root cause”
The TL;DR are the screenshots.
Link to post in the reply.
John Urschel just showed that the growth factor for Gaussian elimination with partial pivoting is ~n^{1/2} with high probability, resolving a long-standing conjecture of Nick Trefethen! https://t.co/FraoYI6yh2
If you want to know whether a system is being driven or at equilibrium, you need its probability current.
Getting this from actual dynamical trajectory data has been the hard part.
To help solve this I've been developing a
Delaunay-Voronoi Fokker-Planck (DVFP) instrument.
DVFP turns scattered, irregular, time-ordered phase-space observations into an empirical Fokker-Planck description.
The distinction is simple:
Geometry tells us what is nearby. Time tells us what happens next, and these are handled properly in this construction.
From here we obtain density, drift, diffusion, and probability current, which is what detailed balance and entropy production are built from.
Paper and DOI: https://t.co/2zkcOs7sWZ
Animation: As the underlying dynamics evolve, the DVFP instrument adapts its local weighting while preserving the temporal ordering of the observations.
Over 10 years ago I downloaded a Flash file, consisting of a video and audio. Unfortunately, SWF is now dead.
I've extracted the video and gotten the source file of the audio, and reconstructed them. The audio is already famous, but I wanted to restore the original meme format.
With all the Navier–Stokes excitement, I almost forgot to say that our tiny billiard ball has made it to @PNASNews — and yes, it can run a universal computer! 🎱💻
Undecidability already lives in dimension two. Today with @Isaacramr__ at PNAS.
https://t.co/VbBHtt5MDu
Are probabilities generated by LLMs "coherent"? Coherent means that, if you get bets on non-exclusive events, you can't be on/against a combination of them and make a sure profit (an arbitrage). Together with calibration, this seems a very basic requirement. In fact, it seems to me that every LLM user interested in probabilistic forecasts for finance should benchmark their candidate models.
Sure enough, the authors of this paper that all LLMs are not generating coherent measures.
https://t.co/CWkqhDMHDM
Want to run an entire Tailscale daemon from memory inside a C2 implant with zero disk artifacts, no kernel drivers, traffic indistinguishable from HTTPS to a CDN, and relay connections from the victim network back through the tailnet.
Now you can. Enjoy!
https://t.co/WmBlShAnLr