Andreas Stöckl

[Gerhard Kürner @gerhardkuerner]

Agent schlägt Mensch!

Die Tests, die Andreas (@stoecklai) mit seinen Agenten durchführt, sind ein tolles Beispiel dafür, was im Tagesgeschäft lästig ist und oft zu viel E-Mail-Verkehr führt. Termine von verschiedenen Personen finden und das über verschiedene Kalender und die Bedienung erfolgt über den Bildschirm und nicht über API-Verbindungen oder ähnliches. Danke Andreas für dieses tolle Beispiel!

Die Anwendung in Unternehmen erfordert natürlich einige zusätzliche Faktoren, die in den Consumer-Produkten nicht enthalten sind und die Teil unserer https://t.co/IRrB9rXM39. Plattform sind.

[Andreas Stöckl @stoecklai]

https://t.co/0I4hTUXfRR

[Gerhard Kürner @gerhardkuerner]

New Chunking Method for RAG-Systems
Enhanced Document Splitting
https://t.co/EKQoQV9EmT

[Andreas Stöckl @stoecklai]

I'm #18052 on MultiOnBeta 👉https://t.co/ueWVVo5c09

[Delip Rao e/σ @deliprao]

AI researcher tuning hyperparameters of their LLM

[Andreas Stöckl @stoecklai]

Intelligent Systems Conference 2023 Konferenz https://t.co/L5868GTkpZ

[Andreas Stöckl @stoecklai]

I just published Large Language Models: A New Frontier in Human Mobility Prediction https://t.co/25GBUTE3ZC

[Andreas Stöckl @stoecklai]

@AndrewYNg A similar 2. hear old paper about the representation of the state of a chess board learned by a GPT Model https://t.co/LUrilvEK4A

[Andreas Stöckl @stoecklai]

Check out the new research paper on Gorilla, a fine-tuned LLAMA-based model that surpasses state-of-the-art LLMs! Read more here: https://t.co/KdMJZaNCI2 and listen to the podcast: https://t.co/A0JqkIN5Ep

[Andreas Stöckl @stoecklai]

Check out this new paper on how large language models can use parallel decoding to reduce latency and improve answer quality! Read the full study and listen to the podcast: https://t.co/KdMJZaNCI2 and https://t.co/A0JqkIN5Ep #LLMs #ParallelDecoding #SoT

[Andreas Stöckl @stoecklai]

I just published The Week in Generative AI from 2023–07–02 https://t.co/1UcTkqjqOi

[Andreas Stöckl @stoecklai]

Check out this paper on GPT-3.5 and GPT-4 by andreasstckl: https://t.co/KdMJZaNCI2 & https://t.co/A0JqkIN5Ep

[Andreas Stöckl @stoecklai]

Language models struggle with long input contexts, according to the new research paper "Lost in the Middle" from andreasstckl. Check out the paper & podcast: https://t.co/KdMJZaNCI2 & https://t.co/A0JqkIN5Ep

[Andreas Stöckl @stoecklai]

Introducing #LONGNET: a Transformer variant that can scale the sequence length to over 1 billion tokens! Check out this latest study by andreasstckl & tune into the podcast: https://t.co/A0JqkIN5Ep #AI #NLP #DeepLearning #Research

[Andreas Stöckl @stoecklai]

Textbooks Are All You Need - new research paper on the impact of high-quality data on LLMs for code generation tasks. Check out the findings by andreasstckl & listen to the podcast: https://t.co/A0JqkIN5Ep #AI #ML #Data #CodeGeneration

[Andreas Stöckl @stoecklai]

Check out this new research paper on AudioPaLM, a multi-modal language model that combines the strengths of text-based and speech-based models! #AI #NLP #LanguageModels #Research https://t.co/KdMJZaNCI2

[Andreas Stöckl @stoecklai]

Check out this podcast to learn more about #CausalNLP #LLMs #CORR2CAUSE: https://t.co/r5jizLCcLi

[Gerhard Kürner @gerhardkuerner]

"Die künstliche Intelligenz wird den Medienmarkt auf den Kopf stellen"

https://t.co/02qW6w0NDW

[Andreas Stöckl @stoecklai]

Challenging prevailing beliefs: Refined web data alone can lead to powerful language models. Check out the findings of this groundbreaking research paper: https://t.co/iOa3UE2QtL #languageModels #webData #research #datascience

[Andreas Stöckl @stoecklai]

Check out my latest #podcast https://t.co/BdXOiV9ywN on #Podbean