As far as I can tell, the only clue that that text‑davinci‑002 wasn't RLHF-tuned was this footnote in the InstructGPT announcement: https://
openai.com/blog/instructi
on-following/#fn2
…
GENERATIVE AI
-
Text-davinci-002 RLHF tuning clue in InstructGPT announcement footnote
By
–
-
Is text-davinci-002 RLHF-tuned? A footnote suggests otherwise.
By
–
As far as I can tell, the only clue that that text‑davinci‑002 wasn't RLHF-tuned was this footnote in the InstructGPT announcement: https://
openai.com/blog/instructi
on-following/#fn2
… -
CICERO AI Diplomacy Champion Discusses Simultaneous Planning and Conversation
By
–
💬 Making plans, conversing simultaneously: 3x Diplomacy World Champ @thegoffy and @adamlerer talk about CICERO’s ability to make plans & converse w/ other players concurrently during gameplay.
— AI at Meta (@AIatMeta) 30 novembre 2022
Watch the full video: https://t.co/Lk4R4xW0jj
Would you play #CICERObyMetaAI? pic.twitter.com/mD5gPigQF8Making plans, conversing simultaneously: 3x Diplomacy World Champ @thegoffy and @adamlerer talk about CICERO’s ability to make plans & converse w/ other players concurrently during gameplay. Watch the full video: https://
youtu.be/kexYmcu1Zro Would you play #CICERObyMetaAI? -
Sparse GPT-3 Models Achieve 3x FLOP Reduction on CS-2
By
–
We trained unstructured sparse 1.3B GPT-3 models on CS-2 systems and demonstrated how we achieve competitive results at a fraction of the inference FLOPs – our 83.8% sparse model achieved a 3x reduction in FLOPs at matching performance Learn more here:
-
Data as Genesis for Modern Innovation in Machine Learning
By
–
“Data is the genesis for modern innovation” says Swami Sivasubramanian, Vice President of AWS Data and Machine Learning at #AWS #reInvent
-
Google Booth: Functional View of Generative Models Discussion
By
–
Drop by the Google booth today at 10:30am to hear @ziwphd discuss a functional view of generative models — generating functions by learning from functions! pic.twitter.com/8cwm4rJhMv
— Google AI (@GoogleAI) 30 novembre 2022Drop by the Google booth today at 10:30am to hear @ziwphd discuss a functional view of generative models — generating functions by learning from functions!
-
Crypto AI Art Collection Reaches 50 ETH Floor Milestone
By
–
Only 15 pieces away from a 50 Eth Floor! Crypto AI Art #023 HISTORICAL SIGNIFICANCE: The "Becoming" Collection is opening the doors to the #CryptoAIArt Movement. https://
opensea.io/assets/ethereu
m/0x495f947276749ce646f68ac8c248420045cb7b5e/2392630434290240917728431095880785304289144848761899072947382465768816902145
… Owned by: sdfdsf Price: PRICELESS #MontrealAI #NFT #NFTCommunity -
BLOOM: 176B Parameter Open-Access Multilingual Language Model
By
–
BLOOM: A 176B-Parameter Open-Access Multilingual Language Model Le Scao et al.: https://
arxiv.org/abs/2211.05100 #ArtificialIntelligence #DeepLearning #MachineLearning -
Text-davinci-002 derived from code-davinci-002 with code training
By
–
News to me that text-davinci-002 derives from code-davinci-002. I imagined the code model had code training the text one lacked. It looks like they both have code training, but instruction tuning interferes with code completion use cases maybe.
-
OpenAI retroactively updates training methods for GPT-3.5 models
By
–
OpenAI update on training methods for “GPT‑3.5” models (retroactively including text‑davinci‑002): https://
beta.openai.com/docs/model-ind
ex-for-researchers
…