GPT-4 Technical Report
Authors:
OpenAI,
Josh Achiam,
Steven Adler,
Sandhini Agarwal,
Lama Ahmad,
Ilge Akkaya,
Florencia Leoni Aleman,
Diogo Almeida,
Janko Altenschmidt,
Sam Altman,
Shyamal Anadkat,
Red Avila,
Igor Babuschkin,
Suchir Balaji,
Valerie Balcom,
Paul Baltescu,
Haiming Bao,
Mohammad Bavarian,
Jeff Belgum,
Irwan Bello,
Jake Berdine,
Gabriel Bernadett-Shapiro,
Christopher Berner,
Lenny Bogdonoff,
Oleg Boiko
, et al. (256 additional authors not shown)
Abstract:
We report the development of GPT-4, a large-scale, multimodal model which can accept image and text inputs and produce text outputs. While less capable than humans in many real-world scenarios, GPT-4 exhibits human-level performance on various professional and academic benchmarks, including passing a simulated bar exam with a score around the top 10% of test takers. GPT-4 is a Transformer-based mo…
▽ More
We report the development of GPT-4, a large-scale, multimodal model which can accept image and text inputs and produce text outputs. While less capable than humans in many real-world scenarios, GPT-4 exhibits human-level performance on various professional and academic benchmarks, including passing a simulated bar exam with a score around the top 10% of test takers. GPT-4 is a Transformer-based model pre-trained to predict the next token in a document. The post-training alignment process results in improved performance on measures of factuality and adherence to desired behavior. A core component of this project was developing infrastructure and optimization methods that behave predictably across a wide range of scales. This allowed us to accurately predict some aspects of GPT-4's performance based on models trained with no more than 1/1,000th the compute of GPT-4.
△ Less
Submitted 4 March, 2024; v1 submitted 15 March, 2023;
originally announced March 2023.
High-quality hexagonal boron nitride from 2D distillation
Authors:
Huanyao Cun,
Zichun Miao,
Adrian Hemmi,
Marcella Iannuzzi,
Jürg Osterwalder,
Michael S. Altman,
Thomas Greber
Abstract:
The production of high-quality two-dimensional (2D) materials is essential for the ultimate performance of single layers and their hybrids. Hexagonal boron nitride (h-BN) is foreseen to become the key 2D hybrid and packaging material since it is insulating, tight, flat, transparent and chemically inert, though it is difficult to attain in ultimate quality. Here, a new scheme is reported for produc…
▽ More
The production of high-quality two-dimensional (2D) materials is essential for the ultimate performance of single layers and their hybrids. Hexagonal boron nitride (h-BN) is foreseen to become the key 2D hybrid and packaging material since it is insulating, tight, flat, transparent and chemically inert, though it is difficult to attain in ultimate quality. Here, a new scheme is reported for producing single layer h-BN that shows higher quality and much more uniformity than material from chemical vapor deposition (CVD). We delaminate CVD h-BN from Rh(111) and transfer it to a clean metal surface. The twisting angle between BN and the new substrate yields metastable moiré structures. Annealing above 1000 K leads to 2D distillation, i.e., catalyst-assisted BN sublimation from the edges of the transferred layer and subsequent condensation into superior quality h-BN. This provides a new and low-cost way of high-quality 2D material production remote from CVD instrumentation.
△ Less
Submitted 24 October, 2020;
originally announced October 2020.
