Self-Improving Pretraining: using post-trained models to pretrain better models Paper • 2601.21343 • Published 8 days ago • 15
Running on Zero Featured 104 Echo-TTS Preview 🐢 104 Fast, multi-speaker TTS (44.1kHz) with voice cloning
view article Article KV Caching Explained: Optimizing Transformer Inference Efficiency Jan 30, 2025 • 230
TAM-Eval: Evaluating LLMs for Automated Unit Test Maintenance Paper • 2601.18241 • Published 11 days ago • 7
PaperBanana: Automating Academic Illustration for AI Scientists Paper • 2601.23265 • Published 7 days ago • 124
THINKSAFE: Self-Generated Safety Alignment for Reasoning Models Paper • 2601.23143 • Published 7 days ago • 38
ASTRA: Automated Synthesis of agentic Trajectories and Reinforcement Arenas Paper • 2601.21558 • Published 8 days ago • 56
Running on Zero Featured 1.27k Qwen3-TTS Demo 🎙 1.27k Transform text into natural-sounding speech with custom voices