Igor Molybog
Igor Molybog has 2 papers in our library, in 2023, most of them on Language models & pretraining; often with Andrew Poulton, Barlas Oguz, Diana Liskovich.
OpenAlex · ORCID · Wikidata · h-index 6 · 26 works · 2,663 citations OpenAlex, 2026-10-14
📄 2 works in our library · 🔗 Cited 7,262 times in our library · ⭐ Paper Stars ranking: #6751
Affiliations
Fields, topics and concepts
📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Arts and Humanities · 2 worksComputer Science · 2 worksArtificial Intelligence · 2 worksLanguage and Linguistics · 2 worksComputer Vision and Pattern Recognition · 1 work
🧭 Topics 1Research subjects and languages the work is about.Language models & pretraining · 1 work
Works by year
2020s
Papers
- Effective Long-Context Scaling of Foundation Models (2023 · 184 citations)
In 2023, strong long-input handling was mostly available through proprietary APIs. Earlier open attempts to extend context were usually judged only by perplexity or synthetic probes, and they often lost quality on short tasks. This Meta paper introduces Llama 2 Long, a family of 7B–70B models built by continuing to… - A Theory on Adam Instability in Large-Scale Machine Learning (2023 · 15 citations)
Very large language models sometimes show sudden loss spikes during training, and the cause has not been explained.
Co-authors here
- Andrew Poulton
- Barlas Oguz
- Diana Liskovich
- Han Fang
- Hejia Zhang
- Jingyu Liu
- Louis Martin
- Madian Khabsa
- Moya Chen
- Naman Goyal
- Peter Albert
- Prajjwal Bhargava
Cited by
Updated
This page in Markdown → https://md.research.elamur.ai/free/en/authors/igor-molybog/