< prev 11 / 12 more >

• 2Switching an LLM's tier changes its "best tool" answer about half the time (modelsagree.com)
  by arephan1 hour ago | 1 comments
• 3Show HN: Proton – A pure Go GUI library with zero CGO and API immunity
  by CzaxTanmay1 hour ago | 2 comments
• 31Show HN: I RL-trained an agent that trains models with RL (for –$1.3k) (github.com)
  by Danau5tin1 hour ago | 5 comments
• 27Beautiful Type Erasure with C++26 Reflection (ryanjk5.github.io)
  by RyanJK51 hour ago | 4 comments
• 2Now, defenders are embracing the prompt injection, too (arstechnica.com)
  by thm1 hour ago | 0 comments
• 5Lenz – A fact-checking API for AI-generated content (lenz.io)
  by kostaj1 hour ago | 3 comments
• 3Show HN: EdgarKit – a real-time JSON API for SEC EDGAR filings (edgarkit.com)
  by mathewarena1 hour ago | 3 comments
• 2The Unexamined Model Is Not Worth Training (bastian.rieck.me)
  by Pseudomanifold1 hour ago | 0 comments
• 3A Trusty Boot Key (Ventoy Alternative), for Bastille Day (codeberg.org)
  by mixmastamyk1 hour ago | 0 comments
• 3We need a PIT Crew for news (werd.io)
  by benwerd1 hour ago | 0 comments

< prev 11 / 12 more >