← All notes

Effective Long-Context Scaling of Foundation Models

Summary

The authors present long-context language models built through continual pretraining from Llama 2. The authors continually pretrain from Llama 2 with longer sequences and a dataset that upsamples long texts, producing models with effective context windows of up to 32,768 tokens. They report gains on regular tasks and larger gains on long-context tasks over Llama 2. With a cost-effective instruction-tuning step that does not need human-annotated long instruction data, the 70B variant can already surpass gpt-3.5-turbo-16k on a suite of long-context tasks. Ablations find that abundant long texts in pretraining are not the key, and that long-context continual pretraining is more efficient than, and similarly effective to, pretraining from scratch with long sequences. [S1]

Claims

Verification

When to apply

Use when asking how Meta extended Llama 2-style models to 32k context, or how continual pretraining compares with training from scratch.

Facts

Sources

License: this note is CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Summary by iinventedlight (AI-assisted). Paper is CC BY 4.0.

License: CC-BY-4.0 · Summary by iinventedlight (AI-assisted). Source: Wenhan Xiong, Jingyu Liu, Igor Molybog, and 18 coauthors, Effective Long-Context Scaling of Foundation Models (CC BY 4.0).

Cite this

Permanent ID: iil:note:xpe5p591

'iinventedlight (AI-assisted)'. 2026. "Effective Long-Context Scaling of Foundation Models." iinventedlight Library, iil:note:xpe5p591, version of 2026-11-12. https://iinventedlight.com/library/notes/2309-16039-effective-long-context-scaling/. License: CC-BY-4.0.

@misc{iil_note_xpe5p591,
  author = {{'iinventedlight (AI-assisted)'}},
  title = {{Effective Long-Context Scaling of Foundation Models}},
  howpublished = {iinventedlight Library},
  year = {2026},
  date = {2026-11-12},
  url = {https://iinventedlight.com/library/notes/2309-16039-effective-long-context-scaling/},
  note = {Permanent ID iil:note:xpe5p591. License: CC-BY-4.0.},
}

Markdown

Spotted a mistake? Use the MCP feedback tool or email hello@iinventedlight.com.

For language models and agents: start at /llms.txt | /agents.md | /.well-known/agent-library.json. This page in Markdown: /library/notes/2309-16039-effective-long-context-scaling.md.