A curated collection of papers and blogs on long-context LLM modeling, with a focus on surveys, efficient attention, KV-cache optimization, memory, RAG, and related topics. The repo was created in 2023 and has activity up to 2026-07-27, with 2155 stars and 101 forks.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
This repository curates papers and blogs on long-context language modeling, covering surveys; efficient attention; KV-cache optimization; recurrent transformers and state-space models; position encoding & length extrapolation; long-context training; long-term memory; retrieval-augmented generation; in-context learning; context and model compression; long reasoning (long CoT); long video & image; long-horizon agents; long-text generation; inference acceleration; benchmarks & evaluation; and technical reports.
How it works
The project aggregates materials related to long-context modeling, including survey papers, efficient attention techniques, KV-cache solutions, memory-centric approaches, RAG, and evaluation resources. The README presents a taxonomy and links to content such as a survey paper, notes, and related repositories.
Getting started
Install/usage instructions are not provided in the truncated README excerpt. The repository is described as an index of papers and blogs rather than a code library.
Recent releases
Releases section indicates the latest release is listed as "latest 0" with: - none.
Traction
Stars: 2155. Forks: 101. Open issues: 0. Created: 2023-09-17. Last push: 2026-07-27. Language: - . License: MIT.
Behind the repo
Not provided in the excerpt.
Caveats
License: MIT. Created: 2023-09-17. Last push: 2026-07-27. Open issues: 0. No explicit runtime or installation requirements are shown in the truncated README.