We are happy to announce that 67 new projects have been awarded grants today as part of the Next Generation Internet intiative, across three different funds: NGI Zero Commons Fund, NGI TALER and NGI Fediversity. We congratulate the developers and engineers involved with these projects, and thank them for their forthcoming contribution to an open, resilient and human-centered internet. The selection covers the entire technology stack from trustworthy open hardware, to services & applications p... Read more ›
After dev kit success, Modos’ cofounders offer an e-paper monitor Read more ›
libffi re-derives the same argument layout on every call. A new API lets a caller build that layout once and reuse it, several times faster, with no writable-executable memory. Read more ›
Offensive content in the eyes of a software engineer. The post appeared first on . Read more ›
The Amazon founder complained to Trump over dinner that nobody there listened to him before slashing staff, according to excerpts from a forthcoming book. Read more ›
AI-assisted software development has moved from line-level autocomplete to agents that can plan changes, edit files, and submit pull requests with limited human supervision. Open-source software, however, evolves through a process designed for humans: contributor agreements, codes of conduct, and review norms all assume a legally accountable person who can attest to provenance and answer reviewer questions. Autonomous and semi-autonomous AI cont... Read more ›
A lightweight programming language built with Go. Contribute to confh/Tiny development by creating an account on GitHub. Read more ›
Today, Chinese AI startup Z.ai (formerly Zhipu AI) , a 753-billion parameter open-weights large language model (LLM) engineered specifically to dominate "long-horizon" autonomous coding and engineering tasks. Available immediately on , the, and more than 20 third-party coding environments, the model boasts a highly stable 1-million-token context window alongside enterprise subscription tiers starting at just $12.60 per month. In excellent news for cost and security-conscious businesses, z.ai ... Read more ›
Maybe there will even be more interest in the invention and use of more advanced algorithms and data structures that use less memory? Read more ›
Vision models can achieve strong performance on classification tasks, but the internal representations supporting their predictions are often difficult to interpret. This work investigates whether sparse autoencoders can decompose intermediate representations of a vision model into interpretable features. We train a ConvNeXt classifier on the FGVC-Aircraft dataset, extract spatial activations from its final feature stage, and train a sparse auto... Read more ›
GoldenEye Recomp v1.0 has been released, providing 'a native PC port of GoldenEye 007 built by statically recompiling the original game into C++' with no emulation involved. Read more ›
Subquadratic is a frontier AI research and infrastructure company building a new class of LLMs. Read more ›
A guide for doing static builds for Haskell projects with Nix. Read more ›
And say hello to much better NTFS support. Linux creator Linus Torvalds has announced the release of Linux 7.1. Read more ›
European public institutions are betting big on digital sovereignty and open source software. And yet they recently moved their Bluesky accounts to W Social, a private, for profit network owned by Swedish entrepreneurs that has quietly gone closed-source. Read more ›
This write-up details a novel iPhone BootROM vulnerability discovered and exploited by our team. It covers the underlying bug, the associated exploitation techniques, and the post-exploitation steps required... Read more ›
I suppose it's been awhile since I wrote a proper announcement. Ah, the Read more ›
Like a cult classic movie, Godot 4 has only gotten better with age. This brings us to Godot 4.7. With 3 years under its belt, the 4.7 _Director’s Cut_ offers colors of never-before-reached intensity. Read more ›
Multimodal large language models (MLLMs) excel at visual reasoning but rely on text-based chain-of-thought (CoT), lacking interpretable visual intermediates. Existing methods use opaque tokens or external tools, missing key properties. We propose Gen-VCoT, a framework using expert vision models to generate RGB images as reasoning intermediates. It has three stages: visual grounding (SAM segmentation), geometric reasoning (Marigold depth maps), a... Read more ›