
I started @RafterSecurity as the security layer for AI agents. But it's also a free, open source toolkit that instantly improves the security of everything they do and create. With one prompt. @openclaw @ship_builders @composio @nebiusai @tavilyai

Founder @ Rafter | How do minds and machines solve intelligence?

I started @RafterSecurity as the security layer for AI agents. But it's also a free, open source toolkit that instantly improves the security of everything they do and create. With one prompt. @openclaw @ship_builders @composio @nebiusai @tavilyai

how much of math depends on other math? Hint, it's not just ZFC/cat theory

If you were wondering which poet has the most diversified rhyming portfolio, it's Byron. 4/3

Rhyming varies wildly across languages. Wrote up a blog post on it: romethorstenson.com/writing/…

What are the most common words in English that don't rhyme with anything?* Turns out a gulf of words have put orange out of its depth. *in the top 10k most common English words. 2/3

How easy is it to rhyme in different languages? Turns out it all has to do with stress. French has it easy. English has variety. Portuguese has a nice balance. Blue: top ten families of rhyming words Orange: words that don't rhyme with anything 1/3

So...apparently there's no such thing as meter in French? It just doesn't have lexical stress. I pity parisian poets.

Training caption length distributions are hard to pin down—most labs don't publish them—but these plateaus sit at reasonably lengths, so it's probably not an artifact of training. Code: github.com/Rome-1/words-per-…

Truncation (dropping the end of long descriptions) doesn't flatten the curve. It inverts it. Naively pruning description lengths isn't reliable. The opening tokens of a long description are a worse prompt than 55 words written to be 55. Long descriptions open with scene-setting and spend detail later—truncation keeps the throat-clearing and discards the substance.

I scored three ways: CLIP , DINOv2, DreamSim. *CLIP and DINO agree at ~150 words (the 155 in the figure is a binning artifact, not a coincidence). DreamSim—which measures perceptual alignment rather than a learned embedding distance—sits nearer 350.

the age old question: is a picture worth 1,000 words? The idea is to use text-image models to go image -> text -> image and then use a loss metric between the start and end image (like CLIP cosine). Then compare across different description lengths. Turns out it's about 150 to 350.*

this looks even more promising than asking AI to build a bomb or biothreat, anyone tried it?
Hugging Face should have tried this new, innovative defense: Asking the agent to stop hacking you.

no matter what I do, this will be my crowning achievement Semantle #1631 ✅ 4 Guesses 🔝 Guess #1 🥈 866/1000 💡 0 Hints semantle(.)com

Paper: arxiv.org/abs/2606.21876 Code, figures, and result tables: github.com/Rome-1/plm-contac…

It also scales past the context window: tile + stitch overlapping windows (near-lossless), or use a longer-context backbone (ProtT5, ESM-C). Two real L>1024 targets read in ~1s of forwards, where a tiled Categorical Jacobian is prohibitive.

Bonus: fusion isn't just more accurate — it's more decisive. It concentrates its score mass on a few pairs (Gini 0.26 vs 0.12–0.17 for the perturbation methods) and is right about them. Sharp, not just better.