We introduce phi-3-mini, a 3.8 billion parameter language model trained on 3.3 trillion
tokens, whose overall performance, as measured by both academic benchmarks and internal …

We introduce phi-1, a new large language model for code, with significantly smaller size than
competing models: phi-1 is a Transformer-based model with 1.3B parameters, trained for 4 …

Most factor-based forecasting models for the term structure of interest rates depend on a
fixed number of factor loading functions that have to be specified in advance. In this study, we …

We present phi-4, a 14-billion parameter language model developed with a training recipe
that is centrally focused on data quality. Unlike most language models, where pre-training is …

… Kauffmann presents us in this volume wvith no- thing less than a new … Kauffmann has
succeeded admirably. His aesthetic … Kauffmann rightly points out, what was really needed after a …

… In ihm erkennen wir einen der beiden verschollenen Engel des Jacopo di Piero Guidi wieder.
Er tritt … Ergänzung zum Oeuvre des Jacopo di Piero Guidi, nämlich als sein frühestes Werk. …

II Maestro della Pala delFaltar maggiore di S. Martino a Mensola (dono della Famiglia fiorentina
Zati nell’anno 1391) viene determinato come tardo seguace di Andrea Orcagna. Le sue …

… Sieben hilner glaubt, dab fur das Palaologus-Portrat der Geifielung dem Piero ein alteres …
Dieser gleiche Typus hatte dann sowohl fur Piero als auch fiir Gozzoli als Quelle zu gelten. …

… Kauffmann presents us in this volume with nothing less than a new … Kauffmann has
succeeded admirably. … Kauffmann rightly points out, what was really needed after a long .period …

… Piero nicht gut: der Verlauf der Ranken wurde gestört (wo rechts oben aus dem obersten Blatt
des Lorenzo di Gio vanni ein Stengel hätte entsprießen müssen, hat Piero … Piero arbeitete …