To clarify: I'm not saying any LLM consistently acts like it believes 9.11 > 9.9 no matter how you prompt it. I'm saying if you prompt in this specific way, many leading models will tell you 9.11 > 9.9 and that's weird. If you want to repro, pls paste prompt from the alt text.
RESEARCH
-
Dolphin 2.9.3-Yi-1.5-34B Fine-Tune Reaches 111k Downloads
By
–
Kudos to @bartowski1182 and @cognitivecompai for the remarkable Yi fine-tune: Dolphin 2.9.3-Yi-1.5-34B-32k-GGUF on @huggingface
! With over 111k downloads last month, your work is a game-changer in the AI community. Check out the model here: -
Comprehensive Review of The Atomic Human and Trustworthy AI Blueprint
By
–
Very comprehensive and generous review of The Atomic Human by @glawton Meet The Atomic Human – a draft blueprint for trustworthy AI
-
Indexify: New AI Data Processing and Indexing Platform
By
–
Announcing Indexify https://
bit.ly/3zD4szc
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
Training stability control with loss and gradient norm thresholding
By
–
Cool! For the spike I'd try e.g. `-sl 7 -sg 7` to keep instability in check earlier in the training. (will skip update if loss/gradnorm > 7 sigma outlier is detected)
-
Computer Simulation as Cost-Effective Alternative to Physical Demonstrations
By
–
Heu si c'etait pour faire un feu d'artifice où y'avait personne on aurait pu faire une simulation informatique, ça aurait coûté moins cher non ? https://t.co/I7SdntIlBy
— ⚡Jessy SEO formation #IA générative #chatGPT 🦊 (@jessyseonoob) 15 juillet 2024Heu si c'etait pour faire un feu d'artifice où y'avait personne on aurait pu faire une simulation informatique, ça aurait coûté moins cher non ?
-
Jensen Huang and Mark Zuckerberg Discuss AI Research Breakthroughs
By
–
We’re excited to announce that our CEO Jensen Huang will be joined by @Meta CEO Mark Zuckerberg to discuss how research enables AI breakthroughs at #SIGGRAPH2024. Attend the discussion to learn about the intersection of #generativeAI and virtual worlds. https://
nvda.ws/3W3nOoI -
Rio Yokota Presents Neural Scaling Laws Research at ICML
By
–
Sign up now to attend to hear Rio Yokota present at the 7th workshop on Neural Scaling Laws at ICML 2024 Rio from the Tokyo Institute of Technology will present research on continual pre-training of open-source models on Japanese text. July 22, 2024, 9:00AM Local Time
-
Q* Strawberry Algorithm Research Findings and Leaked Papers
By
–
El directo ya lo tenéis resubido. Hemos hablado de 2 artículos con filtraciones y 3 papers que apuntarían sobre la línea de investigación del famoso algoritmo Q* (ahora bautizado Strawberry)
-
Generative AI Impact on Productivity Gains and Challenges
By
–
Do you believe that #GenerativeAI will increase your #productivity? This is an impactful read from @mercer, authored by experts @KateBravery and William Self. They delve into research from one of their latest papers, sharing their perspectives on the gains and challenges we are