psrben/visionhope
Official PyTorch implementation of VisionHOPE: Visual Backbones as Self-Modifying Learning Systems.
GITHUB SIGNALS / PROJECT INDEX
Search the archive. Follow your curiosity. Keep the projects you want to come back to.
Official PyTorch implementation of VisionHOPE: Visual Backbones as Self-Modifying Learning Systems.
Evidence-based code review skills for Codex and Claude Code
True single-frame stills from MiniMax H3 in ComfyUI: a one-frame latent and a clean decode, at any size. No extra models.
An unofficial local visualizer for Jev (TypeSafe): a live view of every call your code makes. Not affiliated with TypeSafe AI.
Open the project to explore the idea and the code.
Open the project to explore the idea and the code.
A native GameCube and Wii emulator for jailbroken PS5, powered by Dolphin. Cover-flow library, box art, memory cards, original music. - elripalda/Porpoise-Dolphin-Emulator-for-PS5
Cron job manager + process supervisor in a single binary. Web dashboard, TUI, retries, failure notifications. crond/supervisord replacement
Deterministic security for real-world agentic applications
Open-source skills that empower any AI agent (Claude, Cursor, Hermes, etc.) to generate end-to-end marketing campaigns - UGC videos, ad videos, product photography and more.
Open the project to explore the idea and the code.
Query Zarr stores with SQL directly from DuckDB.
[EMNLP 2026] Official implementation of V-Rubrics: Visual Faithfulness via Rubric-Based Reinforcement Learning.
DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generation
Zelda Wind Waker HD Recomp.
Fan made Snatcher translation patch for Sony Playstation
OscarWatch Multiplatform Satellite Tracking System
A WebAssembly runtime implemented in Erlang/OTP.
Vendor Neutral, Rust-Based Management UI and CLI for U2F/FIDO2 and other hardware security keys - GitHub - framefilter/keyroost: Vendor Neutral, Rust-Based Management UI and CLI for U2F/FIDO2 and other hardware security keys
Anima Detail Enhancer Node for ComfyUI.
Private AI in your browser tab: chat, images, voice and documents, with the models running on your own device. No server, no account.
The set-and-forget LLM engine for Pascal and Volta: PXQ codec + kernels, auto-tuned per card. Ready-to-run PXQ models in MODELS.md; benchmarks vs llama.cpp in the README.
llama.cpp on a jailbroken PS5: run any GGUF model natively on the console, starting with Qwen 3.8
Turn a raw video into a game-ready sprite sheet.