< prev1/10 more >

• 10ZAYA1-8B: An 8B Moe Model with 760M Active Params Matching DeepSeek-R1 on Math (firethering.com)
  by steveharing11 hour ago | 2 comments
• 52ProgramBench: Can Language Models Rebuild Programs from Scratch? (arxiv.org)
  by jonbaer6 hours ago | 9 comments
• 17SingleRide: Longest route on NYC Subway without visiting the same station twice (singleride.nyc)
  by TMWNN1 day ago | 4 comments
• 45Making LLM Training Faster with Unsloth and NVIDIA (unsloth.ai)
  by segmenta3 hours ago | 2 comments
• 21Show HN: Agent-skills-eval – Test whether Agent Skills improve outputs (github.com)
  by darkrishabh4 hours ago | 4 comments
• 310Google Cloud fraud defense, the next evolution of reCAPTCHA (cloud.google.com)
  by unforgivenpasta16 hours ago | 68 comments
• 22Show HN: Trust – Coding Rust like it's 1989 (github.com)
  by wojtczyk4 hours ago | 5 comments
• 260From Supabase to Clerk to Better Auth (blog.val.town)
  by stevekrouse17 hours ago | 44 comments
• 57What I learned making an app for my family (mendelgreenberg.com)
  by chabad36020 hours ago | 9 comments
• 19Show HN: Social Network for Corporate Cringe (CringeOut.com)
  by CringeOut2 hours ago | 14 comments

< prev1/10 more >