My mental model of Sora is that it is the “GPT-2 moment” for video generation. GPT-2, which came out in 2018, could generate paragraphs of text that are coherent and grammatically correct. GPT-2 wasn’t able to write an entire essay without making mistakes like being inconsistent
RESEARCH
-
Consciousness as Pattern: Substrate-Agnostic Identity Construction
By
–
Consciousness itself has no identity, it's a pattern. The traits and memories over which identity is constructed are not easily extracted from the substrate, it's easier to construct a new, substrate agnostic identity
-
Three Training Phases Required for LLM Performance
By
–
Professor Stephen Bach offers an exceptionally clear description of the three training phases required to make LLMs perform. https://
snorkel.ai/large-language
-model-training-three-phases-shape-llm-training/
… #largelanguagemodels #llmtraining #genai -
Optimizing Reconstruction Steps for Better AI Model Performance
By
–
could you try using fewer reconstruction steps? that much actually work better in this case
-
Citing AI Models with Dates for Research Reproducibility
By
–
If you are writing a paper on AI, cite the model number but also the date range in which you used it. These systems are being continually improved and tuned. However, the lack of clear changes or versioning makes replicability hard sometimes (along with inherent LLM randomness)
-
Orange develops specialized AI model for telecom operators
By
–
Le modèle d’IA est entraîné notamment sur des bases de données et un savoir faire telecom propre à Orange. Ce qui permet de proposer une nouvelle solution vraiment innovante à 300 clients professionnels qui sont des opérateurs. Et il y a un même un papier de recherche qui
-
Douglas Adams Right About AGI: Super-Intelligence With Slow Inference
By
–
It would be fitting if the science fiction author most right about AGI turned out to be Douglas Adams. We achieve super-intelligence, but inference is agonizingly slow…
-
Survey Paper Overview: Understanding LLM Architecture and Design
By
–
If you want a resource for learning more about LLMs this survey paper provides a very nice overview. There’s several useful diagrams, including the one below which shows a breakdown of how LLMs are built.
-
U.S. Government AI Executive Order Implementation Progress Report
By
–
How has the U.S. government fared in its efforts to implement the AI Executive Order? @StanfordHAI and RegLab scholars provide an update following the 90-day milestone on agencies' implementation progress and transparency in enacting requirements. https://
stanford.io/3wFFcXu -
Deep Learning Cheatsheet 13-Page PDF Download Available
By
–
Super #DeepLearning Cheatsheet (13-page PDF) by @afshinea and @shervinea => download it here: https://
linkedin.com/feed/update/ur
n:li:activity:7168400362196877312/
…
—————
#NeuralNetworks #AI #Algorithms #MachineLearning #DataScience #Mathematics #DataScientists