Generative AI and its impact! One major catalyst behind the current Generative AI leap is the technology called transformer models. "Its ability to process and code, to compress a lot of information into the neural networks, is very effective.
MACHINE LEARNING
-
LLM Computational Limits and Novel Hash Collisions
By
–
Problems with calculation run deeper than just tokenization though — only so much computation can happen per token. E.g. an LLM could never predict a log of novel hash collisions found via brute force search.
-
AI Models and Retrieval Augmentations Expected in the Legal Sector
By
–
Haven’t seen much specific beyond the bar exam stuff in the news but I would assume models (or at least retrieval augmentations) for law are coming
-
Exploring the Fundamental Limits and Patching Mechanisms of LLMs
By
–
I feel like we need a unified theory of what LLMs can’t do. Most issues (calculation, string ops, grounding) can be patched (via tools, retrieval) so we see them as trivial, not as omens of subtler but similarly fundamental limits. But why is this true?
-
Limits of semantic embeddings in generation
By
–
Unfortunately this is the downside of semantic search and vector embeddings for text generation. It is better at telling you what someone might say next than what someone has precisely said in the past.
-
Shoutout to Ryan Hoque’s Work and Research Paper
By
–
Nice shoutout to @ryan_hoque
’s work in the talk! https://
arxiv.org/abs/2206.14349
v2
… -
Google releases refreshed MC4 corpus and upgraded mT5 checkpoints
By
–
New MC4 refreshed corpus and upgraded mT5 checkpoints from @GoogleAI
. -
Fine-tuned Smaller Models Outperform Larger Foundation Models
By
–
Our research shows that smaller foundation models that are fine-tuned on domain-specific datasets can outperform larger foundation models. We show that a GPT-NeoX 1.4B model that is fine-tuned for 2,000 training steps can perform just as well as the out-of-the-box GPT-J 6B model.
-
Uniphore Partners with Snorkel AI for Conversational AI Acceleration
By
–
@uniphore partners with Snorkel AI to accelerate conversational AI, demonstrate value faster, & shorten customer onboarding time! With Snorkel, Uniphore will reduce the time it takes to label bespoke customer data & train custom #NLP models.
-
Abacus AI offers customizable machine learning and deep learning solutions
By
–
.
@abacusai provides a wide variety of #AI / #MachineLearning / #DeepLearning solutions. And they enable your org to fine-tune these models with your own data to provide even better results: https://
abacus.ai/foundation_mod
els
…
—————
#BigData #DataScience #ComputerVision #NLProc #GenerativeAI