LLMS - AI Dynamics

Massive ChatGPT adoption proves innovation acceleration, says Orange

By

–

14 June 2023 14h05

Pour la CEO de @orange @Cheydema la preuve que des solutions d’#IA génératives #chatGPT a été adopté massivement est la preuve que le besoin d’innovation a encore accélérer.

Les organisations doivent prendre à bras le corps l’innovation au risque de se perdre… #vivatech pic.twitter.com/KRSBDzYRGw
— Defend Intelligence (Anis Ayari) (@DFintelligence) 14 juin 2023

For the CEO of @orange @Cheydema, the proof that generative AI solutions #chatGPT have been massively adopted is the proof that the need for innovation has further accelerated. Organizations must embrace innovation at the risk of getting lost… #vivatech

→ View original post on X — @dfintelligence

14 June 2023

For LeCun, LLMs lack world knowledge

By

@dfintelligence

–

14 June 2023 14h05

Pour @ylecun les LLMs sont aussi limités par leurs connaissances du monde (qu’ils n’ont pas). Si ils sont capable de résoudre des problèmes complexes humains (comme l’examen du barreau) c’est uniquement par leurs capacité à retrouver de l’information. Rien à voir avec une… pic.twitter.com/MhDnp7B9Lk
— Defend Intelligence (Anis Ayari) (@DFintelligence) 14 juin 2023

For @ylecun, LLMs are also limited by their world knowledge (which they don't have). If they are able to solve complex human problems (like the bar exam) it is only through their ability to retrieve information. Nothing to do with a

→ View original post on X — @dfintelligence

14 June 2023

Mistral AI Founded: Guillaume Lample and Team Launch New Venture

By

@arthurmensch

–

14 June 2023 11h50

Totally thrilled to be alongside @GuillaumeLample and @tlacroix6 to create Mistral AI. A lot of work ahead of us!

→ View original post on X — @arthurmensch

14 June 2023

Pass the Philosophy Bac with ChatGPT, 20/20 mandatory

By

@ludosln

–

14 June 2023 11h03

I'm taking the #BacPhilo with #ChatGPT → https://youtu.be/ILaXTpZrcDU 20/20 mandatory #Bac2023 #BacPro2023

→ View original post on X — @ludosln

14 June 2023

GLoRA: Generalized LoRA for Efficient Model Fine-tuning

By

@_akhaliq

–

14 June 2023 5h50

One-for-All: Generalized LoRA for Parameter-Efficient Fine-tuning paper page: https://
huggingface.co/papers/2306.07
967
… present Generalized LoRA (GLoRA), an advanced approach for universal parameter-efficient fine-tuning tasks. Enhancing Low-Rank Adaptation (LoRA), GLoRA employs a generalized

→ View original post on X — @_akhaliq

14 June 2023

WebGLM: Efficient Web-Enhanced Question Answering System

By

@_akhaliq

–

14 June 2023 5h43

WebGLM: Towards An Efficient Web-Enhanced Question Answering System with Human Preferences paper page: https://
huggingface.co/papers/2306.07
906
… present WebGLM, a web-enhanced question-answering system based on the General Language Model (GLM). Its goal is to augment a pre-trained large language

→ View original post on X — @_akhaliq

14 June 2023

Algorithmic Innovation: The Next Frontier Beyond Scaling

By

@alexandr_wang

–

14 June 2023 5h36

biggest questions in ai: – since the transformer, most AI improvement was via scaling up & great engineering
– almost no algorithmic ideas have driven performance, besides RLHF
– future gains will likely require algorithmic innovation how long will they take? who does them?

→ View original post on X — @alexandr_wang

14 June 2023

LangChain SQL chains versus agents: comparative analysis

By

@hwchase17

–

14 June 2023 5h25

we have both in langchain, often for the same purpose! eg SQL chain vs SQL agent pros and cons to each

→ View original post on X — @hwchase17

14 June 2023

Chains vs Agents: Deterministic vs Nondeterministic AI Systems

By

@hwchase17

–

14 June 2023 5h24

assuming im understanding you correctly, this is me is the difference between "chains" and "agents" (at least as far as how I think about them) chains = deterministic sequence of steps = code shell
agents = nondeterministic sequence of steps = llm shell

→ View original post on X — @hwchase17

14 June 2023