antirez

Why I believe there are a lot of misleading papers about LLM reasoning.

That's the whole point. The concept exists almost everywhere but in places like Sicily and China it assumes a fundamental role in life.

Today I discovered that in China there is this concept of "losing face" (mianzi / 面子), which maps 1:1 with Sicilian "perdiri la facci" both as exact wording and conceptually.

YouTube video by Salvatore Sanfilippo Vector databases and recall: exploring the concept in a more deep way.

Saying that a vector database reaches a given recall percentage is more marketing than anything, HNSWs are more complicated than that. Let's explore why in this video. www.youtube.com/watch?v=LG3q...

YouTube video by Updates From YouTube Made On YouTube 2025: Auto-Dubbing

Finally YouTube is going to improve the dubbing feature with lipsync and potentially (looking at this example video) voice matching. That's great in general, and I'm very happy for my YouTube channel that can be better experienced by english speakers. www.youtube.com/watch?v=8W3n...

Who are the designers of the new iPhone UI? Would love to hire them in order to create the hardest escape room in the world.

We have it now! "WITHATTRIBS" :) It's documented in the README, and I'm here for any other info. Merged into unstable some time ago.

VRANGE was merged into Redis unstable and will be likely back-ported. It allows to get lex ranges of elements in vector sets, which is useful as iterator and for other tasks.

It's a few days that I'm writing a blog post on HNSWs. As usually, writing about a subject raises new questions, that trigger new searches, and in the goal of explaining what you understood, you understand more. Blogging may no longer be cool but surely remains useful.

YouTube video by Salvatore Sanfilippo Inspecting LLM embeddings in GGUF format with gguflib.c and Redis

New EN video: Inspecting LLM embeddings in GGUF format with gguflib.c and Redis

www.youtube.com/watch?v=ugba...

ADV: my C course on YouTube is free and can be watched without problems with subtitles or dubbed audio by non Italian speakers :)

LLMs are not just data dumps (with errors) are powerful data manipulation and transformation devices.

This also needs to contain devices designed to last, with simple UIs (keyboards and displays) that can be powered on (solar power maybe?) to ask questions straight away. GPUs may long be gone at some point.

We should likely put a large LLM weights and the explanation of how to do inference in some hard-to-degradate material, inside N bunkers in different parts of the world. So that humans are more likely to be able to restart civilization in case of fatal events.

Still, I believe there is no better way so far in order to archive the human history for the long term. The web (it is now very clear) loses pieces every year. I really hope that the LLMs weights will be saved for a long time as an archive, incorrect very in many details, but better than nothing.

That's however is, in my opinion, a *side effect*. I refute the idea that LLMs are mainly just lossy compression of information: if we don't add to this view their ability to combine and process such information, we are providing a strongly insufficient mental model.

At this stage AI will provide a huge boost to balanced people that don't jump in any extreme hype / anti-hype train, and will create issues to the folks that don't see that judicious application of LLMs is the key. One of those historical moments when wisdom is an advantage.

Playing with adding LLMs embeddings into Redis vector sets, to explore relationships and differences with embedding words models like Word2Vec. Will make a YouTube video and/or blog post.

The problem is there, it's the solution that is totally useless. I'm not denying the problem. But it is self-evident when the PR is a mass of AI generated code. At the same time, AI can help to refine a PR and produce *better* code. Also contributors build trust over time usually.

Why the "declare AI help in your PR" is useless, courtesy of 3000 years ago logic:

- If you can't tell without the disclaimer, it's useless: people will not tell.

- If you can tell, it's useless: you can tell.

Just evaluate quality, not process.

You are not there for some time now. Maybe you can't imagine the level the posts in the timeline have reached: it is TERRIBLE. All full of sexist racist shit I don't follow. Also things that don't make me sleep at night like footages of incidents and alike. What a total trash.

It's worse than that: I'm in both communities because for work concerns I can't afford to cut most of the community of Redis, but there are also people there that kinda questioned the fact I encourage switching here. Also many top accounts (not Musk/Trump supporters) stay there for simplicity.

When you use coding agents for something that produces the bulk of the code of an application that will be used for years, also factor in the technical debt that you are happily accumulating. When you use LLMs an an aid, you could, on the contrary, improve your coding culture.

New blog post: AI is different.

antirez.com/news/155

Experts self distillation, RL on long context directly vs incremental, XML format, exact optimizer parameters, there is quite some meat there.

Are you referring to that *exact* technique where in x' = sin(x) * m, you take a lookup table of four shifts and do (m>>s1) + (m>>s2) + (m>>s3) + (m>>s4)? I never saw this used in the past, AFAIK, and LLMs are not aware of it. Note that it is not a lookup table for the sin(x) value itself.

That's the best explanation I got of a technique I developed during Covid for Z80 (Spectrum) 3D programming.

GPT-5 seems good at code reviews. The first model from OpenAI in a *long* time I'll likely start using constantly for such an activity, together with the others.

LLM evaluation: only trust people with hard problems. Today models perform decently on most trivial task: this is a win of the technology, but also means that evaluation is more and more a realm of sprcislists.

So it looks like Anthropic understood money are on SOTA coding models, and is happy with lagging a bit behind in other benchmarks / performances. Makes quite sense after all, especially if there is a tension fine tuning for different goals.