AI - Artificial Intelligence

• From RLVR to RLSVR: Task Transformation Induces Self-Verifiable Rewards for Open-Ended LLM Self-Improvement 80 upvotes

  Reinforcement Learning with Verifiable Rewards (RLVR) has driven recent progress in reasoning-oriented large language models (LLMs) by enabling large-scale optimization

• LongHorizon-Harness: Advancing Long-Horizon Agents for Real-World Tasks 66 upvotes

  Large language model (LLM) agents increasingly undertake long-horizon tasks that require sustained reasoning, tool use, and revision across many interdependent steps

• Mental World Modeling 64 upvotes

  World models enable a predictive substrate for planning and action, yet existing formulations merely answer a physical question: what/where it is, and how will it evolve

• SwanTale: Unified Multi-Speaker Speech and Audio Generation for Instruct and Zero-Shot Tasks 58 upvotes

  Speech and audio generation is often needed in animation dubbing, audio drama, movies, advertising, games, podcasts, and short-video production

• N_0-VTLA: Scaling Vision-Tactile-Language-Action Model with Latent Tactile Tokens 53 upvotes

  We present N_0-VTLA, a vision-tactile-language-action (VTLA) foundation model capable of (1) fine-grained contact-rich manipulation with tactile perception and tactile-feedback control, and (2) offline policy improvement

• Weak-to-Strong On-Policy Distillation 46 upvotes

  On-policy distillation (OPD), which aligns a student with the teacher's token-level distribution on the student's own rollouts, is an effective paradigm for transferring capabilities across LLMs

• Meshy T2: Fast Native Mesh Generation with Flow Matching 44 upvotes

  Polygonal meshes are the standard surface representation of modern 3D pipelines, and generating high-quality meshes with artist-style topology is essential for film, gaming, and interactive 3D applications

• AISPA: User-Centric System Prompt Auditing for Large Language Model Applications 32 upvotes

  System prompts are instructions configured by developers to govern the behaviors of foundation models in AI applications

• N_0-TWAM: Scaling Tactile-Native World-Action Model for Contact-Rich Manipulation 30 upvotes

  We present N_0-TWAM, a tactile-native world-action model for contact-rich manipulation that predicts both future vision and future contact

• Scaling Properties of Text Conditioning in Visual Generation 28 upvotes

  We study empirical scaling properties for text conditioning in visual generation

• XiaomiMiMo/MiMo-Code MiMo Code: Where Models and Agents Co-Evolve TypeScript 12.6k stars
• StarTrail-org/PixelRAG The end of web parsing. The beginning of scalable pixel-native search. link: https://pixelrag.ai/ Python 9.0k stars
• omnigent-ai/omnigent Omnigent is an open-source AI agent framework and meta-harness: orchestrate Claude Code, Codex, Cursor, Pi, and custom agents — swap harness Python 8.1k stars
• elder-plinius/T3MP3ST autonomous red teaming platform; multi-agent offensive-security meta-harness TypeScript 5.4k stars
• EverMind-AI/Raven The memory-first, self-improving agent harness built on EverOS, with MiroThinker-powered deep research and reasoning. Python 3.5k stars
• UditAkhourii/adhd ADHD — a skill for coding agents. Tree-of-thought with pruning, built on the Claude & Codex Agent SDK. Fans out parallel divergent thoughts TypeScript 3.2k stars
• code-yeongyu/lazycodex The one and only agent harness for complex codebases. Project memory, planning, execution, and verified completion inside Codex. TypeScript 3.1k stars
• synthetic-sciences/openscience The open-source AI workbench for scientific research TypeScript 3.1k stars
• TestSprite/testsprite-cli Official TestSprite CLI — AI-powered automated testing from your terminal TypeScript 2.8k stars
• rpamis/comet Comet: agent skill harness for turning ideas into evaluated workflows JavaScript 2.6k stars

Running models on hardware you control: local LLM and agent tooling reads land in Learn under the AI topic.