m_chirculescu's profile picture. You can just do things. DMs open.

Mihai Chirculescu

@m_chirculescu

You can just do things. DMs open.

Pinned

🚀 Launching backtrack_sampler - for experimenting with custom LLM sampling strategies that can rollback and regenerate tokens. It works with both @ggerganov 's llama.cpp (GGUF files) and @huggingface 's transformers github.com/Mihaiii/backtr…


ChatGPT Atlas was released a few hours ago, and already my feed has no posts mentioning it. Is X’s algo tweaked to hype it down? 😄


Free, easy to use, reliable UI automations 🚀

UiPath ScreenPlay 🎬—our fully autonomous, self-driving engine for building and running UI automations—is now FREE for all enterprise and community users.



Mihai Chirculescu reposted

It's a strange time to be a programmer—easier than ever to get started, but easier to let AI steer you into frustration. We've got an antidote that we've been using ourselves with 1000 preview users for the last year: "solveit" Now you can join us.🧵 answer.ai/posts/2025-10-…


Mihai Chirculescu reposted

The latest article in my ACM Queue series on Automind/Universalis for verifiable safe agentic programming. The unwritten conclusion from this last paper is that design by contract pre/post-conditions are no silver bullet. Instead, for safety, we need something to reason about…

Guardians of the Agents Formal verification of AI workflows This proof-based approach addresses problems by providing deterministic and verifiable assurances of safety without the need to trust the AI nor any of the artifacts it produces @headinthebox queue.acm.org/detail.cfm?id=…



Mihai Chirculescu reposted

so... this is my first app teaser hope you like it and sign up for the waitlist at theadhdcopilot dot com


My friend’s hiring a software developer - fully remote, european timezone. Greenfield project. Must have experience with AI tools & building MVPs. JD: drive.google.com/file/d/1Ie1uao… Contact: linkedin.com/in/floriandami…


A new version of Ovis was released 🤩🤩🤩

Introducing Ovis 2.5 - our latest multimodal LLM breakthrough! Featuring enhanced visual perception & reasoning capabilities, consistently outperforming larger models across key benchmarks. #Ovis #MLLM

AI_AlibabaInt's tweet image. Introducing Ovis 2.5 - our latest multimodal LLM breakthrough!
Featuring enhanced visual perception & reasoning capabilities, consistently outperforming larger models across key benchmarks.
#Ovis #MLLM


Mihai Chirculescu reposted

Introducing Ovis 2.5 - our latest multimodal LLM breakthrough! Featuring enhanced visual perception & reasoning capabilities, consistently outperforming larger models across key benchmarks. #Ovis #MLLM

AI_AlibabaInt's tweet image. Introducing Ovis 2.5 - our latest multimodal LLM breakthrough!
Featuring enhanced visual perception & reasoning capabilities, consistently outperforming larger models across key benchmarks.
#Ovis #MLLM

Programming has already changed, anon. You just haven’t been paying full attention.

I never used git worktrees until I started using Claude Code and now I want to use them all the time



That sounds great and I'm hyped. But why not release version 1 now, and version 2 later? It's been four months and there's still no new open-source model. Frankly, this just feels like stalling.

we are going to take a little more time with our open-weights model, i.e. expect it later this summer but not june. our research team did something unexpected and quite amazing and we think it will be very very worth the wait, but needs a bit longer.



There's also ms-swift, which is surprisingly easy to use. For some reason, it's unknown to "westerners" :) github.com/modelscope/ms-…

This is currently the best (and afaik, the only) library that works well with Qwen2.5-VL models. reducto/RolmOCR was trained using LLaMA-Factory. It's still more or less buggy, and most of the discussions/solutions are still in Chinese.

hu_yifei's tweet image. This is currently the best (and afaik, the only) library that works well with Qwen2.5-VL models.
reducto/RolmOCR was trained using LLaMA-Factory.
It's still more or less buggy, and most of the discussions/solutions are still in Chinese.


Mihai Chirculescu reposted

homecooked finetuning tastes better


Apparently, books like "Drive" or "Zero to One" are meant to be skimmed first and only then, if you really feel you need to, read word by word. Is this common knowledge in the U.S.?


Mihai Chirculescu reposted

VLA but smol


🤔🤔🤔🤔OpenAI could easily add a line of code to replace em dashes, but they want to know which text not to train on in the future. 🤔🤔🤔🤔


Mobile devs, what's the equivalent of a hotkey on mobile? @grok says swipe gesture patterns as 'hotkeys' aren't supported on iOS :/ So how do you trigger actions on mobile similarly to how AutoHotKey or AutoIt works on Windows?


My hot take is that it's so much harder to become an indie hacker now than it was a few years ago. That window of opportunity has passed. You're better off deep learning what you believe will still be relevant in the future (in my case, that's the llama.cpp codebase).

If you're interested in doing the nomad coder side gig thing, then listen to this man please!:



Mihai Chirculescu reposted

return to vibe


Mihai Chirculescu reposted

Why Singapore Elects the Losers of its Elections nebula.tv/videos/polymat…

PolyMatters's tweet image. Why Singapore Elects the Losers of its Elections nebula.tv/videos/polymat…

A coding agent that slows down while coding and live streams the code it changes and explains why via audio.


Loading...

Something went wrong.


Something went wrong.