Yüce Tekol

[Georgi Gerganov @ggerganov]

llama.cpp now has an official website: https://t.co/vztdUpdBWL

Our goal is to make local AI accessible to everyone, and improving the user experience is a big part of that. On the new landing page you’ll find a single-line cross-platform installer. The installation provides a single unified `llama` entrypoint which you can use to run/serve models and interface with 3rd-party agentic applications.

While oriented towards simplified user experience, the new `llama` application also provides all the advanced functionality of the existing llama.cpp tooling with which experienced users are already familiar. Also note that all GGUF models that you might have already downloaded with llama.cpp in the past will be automatically available to use without downloading again (they are stored in the common HF cache on your machine).

We have many improvements in the pipeline both at the UX and at the engine level and we plan to iteratively ship new things over the coming months. One of the main focuses will be seamless integration with local-friendly 3rd-party agents (such as Pi). In the meantime, we’ll continue to listen for feedback from the community and adjust accordingly, so keep letting us know what you think and need.

[Yüce Tekol @tklx]

@hannahluci Nice article. But claiming that KK is doing what he does solely because of his ego is a bit naive :)

[Kyle 🚄 @KyleTrainEmoji]

PICARD: Data, shields up

DATA: Brilliant! Shields can reduce damage we sustain. Not immunity. Not hubris. Just prudence. It's not precaution—it's strategy.

[camera shakes]

WORF: HULL BREACHES ON NINE DECKS

DATA: Here's what happened: you told me to raise shields, and I didn't

[Yüce Tekol @tklx]

@malikwas1f That's my experience as well. I always turn thinking off when I run Qwen 3.5/3.6.

[ゆっくりモンド@歴史動画 @Yukkuri_mond]

古代ローマではある時、奴隷であることがわかるように彼らに固有の色の服を着用するよう義務付けようという法案が元老院で審議された

だが「彼らに自分たちがどれだけ数が多いかを思い起こさせるのはまずい」という理由で廃案になったという

[Özgür Özel @eczozgurozel]

Milletimize…

[Chris Avellone @ChrisAvellone]

Fun podcast discussing literary influences on the Fallout universe, from Wasteland to Fallout 1, 2, and New Vegas. : )

Link:https://t.co/lOLX6Euat2

[Yüce Tekol @tklx]

@alexellisuk Are you using any MCP or skills? I didn't try it for code reviews, but `gopls mcp` works great during code generation,

[Yüce Tekol @tklx]

@nero23101 @0xSero You can easily run a 4bit quant of Qwen 3.5 35B with it. I run it on a 4070 ti super ad get around 60 tg/s. It's pretty good for basic/regular coding tasks. I usually use Bartowski or AesSedai quants these days.

[Sam Rose @samwhoo]

I think this is the best post I've ever made.

[IEEE Computational Intelligence Society @ieeecis]

🚨 𝗗𝗲𝗮𝗱𝗹𝗶𝗻𝗲 𝗔𝗽𝗽𝗿𝗼𝗮𝗰𝗵𝗶𝗻𝗴: 𝟮𝟬𝟮𝟲 𝗜𝗘𝗘𝗘 𝗖𝗜𝗦 𝗚𝗿𝗮𝗱𝘂𝗮𝘁𝗲 𝗦𝘁𝘂𝗱𝗲𝗻𝘁 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗚𝗿𝗮𝗻𝘁𝘀!
🗓️ 𝗗𝗲𝗮𝗱𝗹𝗶𝗻𝗲: 30 March 2026
🔗 𝗟𝗲𝗮𝗿𝗻 𝗺𝗼𝗿𝗲 & 𝗮𝗽��𝗹𝘆 𝘁𝗼𝗱𝗮𝘆: https://t.co/4UxNRwC9FB

[Yüce Tekol @tklx]

@relizarov The provider of the quants sometimes makes a big difference. The AesSedai ones worked the best for me for Qwen 3.5 models. Even with that Qwen 3.5 models don't behave predictable for me. They seem to fall into reasoning loops too often.

[IEEE Computational Intelligence Society @ieeecis]

𝟮𝟬𝟮𝟲 𝗚𝗿𝗮𝗱𝘂𝗮𝘁𝗲 𝗦𝘁𝘂𝗱𝗲𝗻𝘁 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗚𝗿𝗮𝗻𝘁𝘀 �� 𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀 𝗡𝗼𝘄 𝗢𝗽𝗲𝗻!
💰 $1,000–$4,000
🎓 IEEE CIS Student Members
⏳ Deadline: 30 March 2026
https://t.co/erbYKGDY7e

[Yüce Tekol @tklx]

@sinanhacir Şimdiki ABD yönetimi İsrail'e tamamen teslim olmuş gibi görünüyor. Üst düzey Cumhuriyetçilerin çoğu da bu durumda olmalı. Ben Rand Paul dışında farklı bir ses duymadım. Öyle olmasa seçimlere bu kadar az zaman varken İran'a da saldırmazlardı.

[Ahmet YAVUZ @yvzah]

Laikliği savunmak suç değil, anayasal bir görevdir.

[Isik Mater @isik5]

You told an LLM "confirm before acting" and expected that to function as a safety mechanism? That's not how these systems work? Prompt-level instructions are not hard constraints.
I wouldn’t think twice if this mistake came from an end user but you work as Director of Safety and Alignment at Meta. You should know that LLM agents misinterpret instructions and over-execute all the time. Sorry but I find this extremely irresponsible.

[Chris Avellone @ChrisAvellone]

.@blockheadcc asked how Annah from Planescape: Torment came to be and the design + aesthetic steps involved, so here’s my answers - which also apply to companion design in general.

https://t.co/nj2dqDgnOD

[MMitchell @mmitchell_ai]

🤖 Pro-tip! Yes, LLMs are appropriately referred to as "stochastic parrots" 🦜. No, entire commercial AI systems are generally not stochastic parrots*. Hope that helps!

[StripMallGuy @realEstateTrent]

The most fascinating documentary I’ve ever seen is called The Up Series.

It followed the lives of a group of 7-year-olds - and filmed them at age 7, 14, 21, 28, 35, 42, 49, 56, and 63.

Biggest takeaway:

Who you are at 7 is likely who you are at 63.