HN · GitHub Signals

OPEN SOURCE, IN FOCUS

weightless

msuiche/weightless

Abliteration without the weights — projective refusal steering as a GLP vector + fail-closed boot hotfix. - GitHub - msuiche/weightless: Abliteration without the weights — projective refusal steering as a GLP vector + fail-closed boot hotfix.

Explore on GitHub ↗

Saved projects stay in this browser. No account needed.

TAKE A CLOSER LOOK

Start with the source.

The repository is the place to check installation steps, supported platforms, the license, and current maintenance before choosing a tool.

Read the documentation on GitHub ↗

Keep exploring

Find projects in the same areas, or search the archive for a different approach.

ANOTHER GOOD FIND

Explore related projects

View the archive ↗

Connected by topics and descriptions.

coleam00

ai-software-factory

A repository that ships without anyone reading the diff: GitHub issues in, merged pull requests out, with the gates that make that defensible.

Take a look →
nicepkg

ctxport

Copy AI conversations as clean Markdown Context Bundles — one click from ChatGPT, Claude, Gemini, DeepSeek, Grok & GitHub. Copy from sidebar without opening. Zero upload, 100% local, open source.

Take a look →
usemoss

moss

The retrieval layer for production AI systems. Lightning-fast (<10ms) search without vector databases. Built for browser, edge, on-device, and cloud.

Take a look →
0xbakeer

qwen3.8-27b-fp8-on-a-single-dgx-spark

Serving Qwen3.8-27B-FP8 on a single DGX Spark (GB10): 7.88 to 58.5 tok/s single-stream from decode strategy alone, weights untouched. Speculative decoding and prefix caching benchmarked, plus DFlash 2 — the only Qwen3.8-27B build that can serve it under vLLM.

Take a look →