Hacker News Reader

linolevanon How we measured AI writing across arXiv, and where the measurement breaks

I’m very skeptical of these results. I run a pretty large group paper website on top of arXiv and we run pangram on papers. The numbers are not nearly this high.

One thing I see a lot is papers flagged as AI because they include llm rollouts in the paper as examples.

linolevanon SQLite improving performance with pre-sort

Neat optimization, I've never seen that before. Thanks for sharing.

linolevanon Knowledge Distillation of Black-Box Large Language Models (2024)

Can we note that this is a 2024 paper in the title?

linolevanon At least 25 Flock cameras have been destroyed in five states since April 2025

Pangram is legit. I don't work at pangram, we integrated it in our paper website and one of the cool emergent behaviors I've seen is that on AI papers with example rollouts, it will accurately mark the paper's main text as human generated and the rollouts as AI generated.

My understanding is that they strongly believe in no false positives, so it's definitely possible to slip something by them but if it marks something as AI, it very likely is.

linolevanon April 2026 TLDR Setup for Ollama and Gemma 4 26B on a Mac mini

What I really don't get is why more people don't talk about LMStudio, I switched to it months ago and it seems like a straight upgrade.

linolevanon Google releases Gemma 4 open models

Hosted on Parasail + Google (both for free, as of now) themselves, probably would give those a shot

linolevanon Qwen3.6-Plus: Towards real world agents

I’m surprised that people are surprised. Qwen has been hosting private plus and max variants for a while now.

linolevanon Subscription bombing and how to mitigate it

Well written piece on an attack vector I'd never thought too hard about before. Thanks for elaborating on why sending an email or two to a random person helps an attacker achieve their goal. A lot of similar articles skip over details like that.

linolevanon Reports of code's death are greatly exaggerated

How much is OpenAI paying you for this

linolevanon Show HN: We built a terminal-only Bluesky / AT Proto client written in Fortran

https://tangled.org/ <--- GitHub on ATProto

that's all I'm aware of

(edit) Oops, just saw that you mentioned it, confused by your first line then. Tangled is awesome!

linolevanon Marvin Hagemeister, Luca Casonato, David Sherret and Phil Hawksworth left Deno

John too:

https://bsky.app/profile/johndonmoyer.com/post/3mhbkepa7gs2v

linolevanon Starlink Mini as a failover

Where on earth are you living with that kind of price point? Unreal.

linolevanon A Recursive Algorithm to Render Signed Distance Fields

Played around with the code to implement a little bit of SIMD. Was able to squeeze out a decent improvement, ~250 fps avg, ~140 low, ~333 high (on an m4). Looks pretty straightforward to do threading with as well. Cool stuff! Could work to bring more gpu stuff back down to the cpu.

linolevanon Claude Doubles Usage Limits During Off-Peak Hours (March 13–27, 2026)

Oops! Looks like we posted at the same time.

linolevanon Claude March 2026 usage promotion

> Does bonus usage count against my weekly usage limit?

> No. The additional usage you get during off-peak hours doesn’t count toward any weekly usage limits on your plan.

linolevanon Shipping Grayscale Photos at Small Scale

Did… you copy paste this from another discussion? I’ve read this comment before.

linolevanon Launch HN: IonRouter (YC W26) – High-throughput, low-cost inference

According to the providers that I keep track of, Cumulus is typically pretty price competitive, except for MiniMax where DeepInfra and Together are much cheaper and GLM-5 where DeepInfra and z.AI's own hosting is much cheaper.

(Also technically qwen3 8b w/ novita being first place but barely)

linolevanon Launch HN: IonRouter (YC W26) – High-throughput, low-cost inference

Can we get context length / output length docs (looks like you mention "Max tokens (chat)" of 128k but it's unclear what that means)? Also it looks like your docs page is out of date from your playground page.

Also piece of feedback: it kind of sucks to have glm/minimax/kimi on separate api endpoints. I assume it's a game you play to get lower latency on routing for popular models but from a consumer perspective it's not great.

linolevanon Sarvam 105B, the first competitive Indian open source LLM

Looks like at least a second release, they had one other LLM before this.

linolevanon Workers who love ‘synergizing paradigms’ might be bad at their jobs

This is awesome. Almost all of these are believable even if you're looking at pretty carefully. I need this on a firestick or something.

linolevanon Nobody ever got fired for using a struct

This is awesome. Got completely lost reading this and was struggling to figure out where I got this link from. Amazing story.

linolevanon The IRIX 6.5.7M (sgi) source code

Yep, I recall one of the big components being libc i18n

linolevanon NanoGPT Slowrun: Language Modeling with Limited Data, Infinite Compute

There was this very interesting paper out of Stanford this last September about pretraining under the unlimited compute but limited data paradigm[0]. Pretty much exactly the same thing but with ~200M training tokens instead.

[0] https://www.alphaxiv.org/abs/2509.14786

linolevanon A case for Go as the best language for AI agents

Not understanding the difference between this and something like cargo audit[0]. I suppose it has something to do with "static analysis of vulnerabilities" but I don't see any of that from a quick google search of govulncheck.

[0]https://crates.io/crates/cargo-audit

linolevanon SSH Snakes.run

Oops. A better link would have been https://eieio.games/blog/secure-massively-multiplayer-snake/.

linolevanon ECS Survivors Parts VII – X

Would recommend that folks start with the original post for context: https://blog.ptidej.net/ecs-survivors-part-i/

linolevanon ECS Survivors Parts VII – X

Thought this was AWS ECS – turns out it was much more interesting! I love entity component systems conceptually, glad to see data oriented design being discussed (even if it's not necessarily positive)

linolevanon Cosmologically Unique IDs

Why? Sharding on lower bits seems fine to me.