Enjoyed this experiment — 128K context is great but you don’t get proportionally more attention. Each output token can only consider so much. Translation — fine
Extraction, doc Q&A — maybe?
Parity bit of input — absurd; the model is not a god
TECHNOLOGY
-
128K context: no proportional attention increase
By
–
-
AI and EdgeAI Shape Canada’s Hardware Ecosystem Future
By
–
Join us today at @ventureLABca
's third-annnual #HardTechSummit! Our CEO, Arun Iyengar, will be sharing insights on how #AI and Hardware EdgeAI shape the future of Canada’s Hardware Ecosystem, alongside executives from at @BetaKit
, @VectorInst
, and @intel
. https://
bit.ly/3sJjWy8 -
Hands-On Experience Essential for Building Complex Systems
By
–
Building rockets, turbojets, neural nets, nuclear power plants, aircrafts, microprocessors,… all of these things require hands-on experience and "black art" knowledge that you can't find written down anywhere.
-
Lowering 3D World Entry Barriers for Mass Adoption
By
–
Right. It will make the barrier to entry really really low, and finally more people will be able to contribute and exist in the 3D world
-
Unreleased AI Innovations Set to Astonish the World
By
–
Agreed, we are seeing some things coming out now that have been under works for some time, but I can imagine that the things being worked on right now that still have not been released will blow us all away
-
Honoring Engineers Scientists Mathematicians Shaping World Today
By
–
Today, we're honoring the engineers, scientists, and mathematicians who are shaping the world! #NationalSTEMDay #STEMDay See what they're doing http://
spr.ly/6018uqGva -
Dell Tech Forum 2023: Innovation and Technology Solutions in Paris
By
–
A la recherche d'inspiration et d'innovations #Tech et #IT ? Venez au #DellTechForum le 23/11 à Paris ! Et si vous ne pouvez pas être sur place, l'événement sera aussi à suivre en livestream Plus d'informations et inscriptions : https://
dell.com/dt/forum-fr?dg
c=sm&cid=dtf23_sm_EMEA_FR&lid=sm_ChoYou_Influenceur&RefID=sm_EMEA_FR_sm_ChoYou_Influenceur
… #innovation #Sponsorisé -
OtterHD: High-Resolution Multi-modality AI Model
By
–
OtterHD: A High-Resolution Multi-modality Model Li et al.: https://
arxiv.org/abs/2311.04219 #ArtificialIntelligence #DeepLearning #MachineLearning -
RISC-V and AI IP at Supercomputing Conference 2026
By
–
We're excited to be in Denver next week at the @Supercomputing Conference to talk about #riscv, AI IP and high density, scalable compute. Come find us at Booth 1707 + if you already know you want to chat drop us a line —> http://
bit.ly/tt_sc23 #AI #supercomputer #IP -
GPT-4 Turbo: Biggest ChatGPT Update Yet
By
–
Biggest ChatGPT Update Yet: GPT-4 Turbo has landed! Get the full story https://
godofprompt.ai/blog/gpt-4-tur
bo-the-biggest-update-for-chatgpt
… Check out these game-changing features: – 128k context window – Updated to April 2023 – 'GPT Builder' for custom chatbots – Multimodal AI for visual learners –