RFC R-State — Resumable State Transfer Format for Heterogeneous LLM Pipeline Architectures
R-State is a proposed serialization format for transferring the internal state of large language models between stages of a heterogeneous inference pipeline without detokenization and re-tokenization. It is designed to reduce latency, preserve intermediate activations and KV-cache, and enable more modular multi-model workflows through a structured packet containing metadata and tensor payloads
Authors
- Pavel Scurin (ORCID: https://orcid.org/0009-0002-5299-4783)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-05-01
- DOI
- https://doi.org/10.5281/zenodo.19951913
- Citations
- 2
- Primary Topic
- Machine Learning and Algorithms
- Type
- article
- Field-Weighted Citation Impact
- 65.68