Our Research
About
apergo.ai is an independent AI research lab aimed at helping to crack the most fundamental challenges in AI. It was initiated and is led by Jörg Frohberg. He is a serial tec entrepreneur, investor and consultant who holds a masters degree in philosophy from Humboldt University and was visiting scholar at Rutgers University.
With the Institute for Applied Informatics at Leipzig University (InfAI) we developed the CRASS benchmark. Currently, we are working with the Lancaster University on a research project analyzing the effect of machine learning in leadership and coaching. Additionally, we participated in Big Science. The BigScience initiative is a research collaboration led by Hugging Face to develop an +100B parameter open source model and a large new data set. We contributed to the creation of the data set, the model and helped draft the data provider framework and agreement.
List of research:
- CRASS: A Novel Data Set and Benchmark to Test Counterfactual Reasoning of Large Language Models accepted at LREC 2022
- Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models
- BLOOM: A 176B-Parameter Open-Access Multilingual Language Model
- The BigScience ROOTS Corpus: A 1.6TB Composite Multilingual Dataset accepted at NeurIPS 2022
- Data Governance in the Age of Large-Scale Data-Driven Language Technology accepted at FAccT ’22
