MacMusic  |  PcMusic  |  440 Software  |  440 Forums  |  440TV  |  Zicos
models
Search

Meta trials Purple Llama project for AI developers to test safety risks in models

Friday December 8, 2023. 12:37 AM , from TheRegister
Security boosted and inappropriate content blocked in large language models
Meta has launched Purple Llama – a project aimed at building open source tools to help developers assess and improve trust and safety in their generative AI models before deployment.…
Read more at TheRegister
https://go.theregister.com/feed/www.theregister.com/2023/12/07/meta_purple_llama_project/

Related News

models Meta's News Ban In Canada Remains As Online News Act Goes Into Effect SlashdotDec 19
developers Oversight Board Criticizes Meta’s Automated Moderation in Israel-Hamas War Wired: Tech.Dec 19
meta AI, efficiency and a rise in citizen developers -- development predictions for 2024 BetaNewsDec 19
project HMRC launches £500M procurement for new ERP, though project's already a 'red' risk TheRegisterDec 19
safety Does Meta's New Face Camera Herald a New Age of Surveillance? Or Distraction... SlashdotDec 18
purple Microsoft Releases Phi-2, a Small LLM That Outperforms Llama 2 and Mistral 7B SlashdotDec 16
llama Open to attack: The risks of open-source software attacks BetaNewsDec 15
trials Mobile app developers turn to AI for testing BetaNewsDec 15
risks Apple expands Self Service Repair to iPhone 15 and M2 Mac models BetaNewsDec 15
models GM's Cruise sheds nine execs in the name of safety and integrity TheRegisterDec 14
developers GM's Cruise Robotaxi Unit Dismisses Nine Execs After Safety Probe SlashdotDec 14
meta Xbox Cloud Gaming Is Now Available On Meta's Quest VR Headsets SlashdotDec 14
project GM Says It's Ditching Apple CarPlay, Android Auto For Your Safety SlashdotDec 13
safety Tesla to remote patch 2M vehicles after damning Autopilot safety probe TheRegisterDec 13
purple OpenPGP for application developers LWN.netDec 13
llama [$] Project Bluefin: A customized Fedora Silverblue desktop image LWN.netDec 12
trials New leader of "unsalvageable" Project Veritas quits, citing "strong evidence" of past illegality BoingBoingDec 12
risks Mistral Says Mixtral, Its New Open Source LLM, Matches or Outperforms Llama 2 70B and GPT3.5 on Most Benchmarks SlashdotDec 11
models Beyond phishing: The top employee security risks you're probably not measuring BetaNewsDec 10
developers Harvard Accused of Bowing to Meta By Ousted Disinformation Scholar in Whistleblower Complaint SlashdotDec 10
News copyright owned by their original publishers | Copyright © 2004 - 2024 Zicos / 440Network
Current Date
May, Thu 9 - 01:11 CEST