Asymmetric Quantization: Near-Lossless Retrieval with 97% Storage Reduction

4 comments

• Ameo 1 hour ago
  I can't wait until we get to 100% storage/cost/compute reduction for LLMs. Every thought you could have thought pre-conceived in high-fidelity super-resolution. Every action you could have taken predicted and simulated in advance courtesy of Openthropic and the USA Sovereign Wealth Fund.
  [-]
  • throwaw12 28 minutes ago
    100% reduction is impossible for something which should work, because -100% means it is now 0
    [-]
    • neonstatic 21 minutes ago
      They were clearly being sarcastic
  • peheje 27 minutes ago
    Reminds me of 'Learning to be me' by Greg Egan
• johnathan101 2 hours ago
  97% is impressive, but I'm curious what the latency tradeoff looks like in production. Storage is only half the story for retrieval systems.
  [-]
• mv_d5339e31 1 hour ago
  [dead]
• TradingReality 1 day ago
  [flagged]