ProRL V2 - Prolonged Training Validates RL Scaling Laws
Prolonged Training Validates RL Scaling Laws
ProMedical: Hierarchical Fine-Grained Criteria Modeling for Medical LLM Alignment (April 2026)
Hierarchical fine-grained criteria modeling for medical LLM alignment — structured clinical evaluation rubrics with multi-level criteria decomposition for improved medical reasoning and safety
prompt
AI prompt for CFO-level financial strategy analysis
`rp2040-pac`
Peripheral access API for the RP2040 dual-core system-on-chip (generated using svd2rust) - [](https://crates.io/crates/rp2040-pac)
Quick Creator
SEO-Optimized Blog platform powered by AI.
prompt
Scientific database orchestrator and molecular research agent system prompt
Qwen2.5-Max
Exploring the Intelligence of Large-scale MoE Model.
Read AI
An AI copilot for wherever you work, making your meetings, emails, and messages more productive with summaries, content discovery, and recommendations.
reactos/reactos
A free Windows-compatible Operating System
prompt
Design LLM evaluation benchmarks like an architect
review-implementing
Evaluate code implementation plans and align with specs.
Scrapling
🕷️ An adaptive Web Scraping framework that handles everything from a single request to a full-scale crawl!
Quill
Write blogs like a human
SanderMertens/flecs
A fast entity component system (ECS) for C & C++
pskill9/web-search
无需 API Key，使用 Google 搜索结果进行免费网页搜索的 MCP 服务器。
proot-me/proot
chroot, mount --bind, and binfmt_misc without privilege/setup for Linux
QuillBot
AI-powered paraphrasing tool.
Rememberizer AI
与 Rememberizer 数据源交互，促进增强的知识检索。
Real-Time C++
companion bare-metal code to Real-Time C++ book.
PySceneDetect
Python and OpenCV-based scene cut/transition detection program & library.
RandyGaul/cute_headers
Collection of cross-platform one-file C/C++ libraries with no dependencies, primarily used for games
replicate/replicate
Discover and run AI models via API
Reasoning Theater: Disentangling Model Beliefs from CoT (2026)
For simple tasks, the model's final answer is already decodable from early-layer activations before CoT generates a single token — CoT produces genuine belief change only on hard problems; probe-guided early-exit reduces token generation by 80% on simple tasks
Quiver
Dump your brain into your GenerativeAI Vault 