Weighted State Souping for Efficient Long-Context Retrieval
Weighted State Souping for Efficient Long-Context Retrieval
Weighted State Souping proposes query dependency weighting to combine cached State state encodings, significantly improving long-context performance on tasks such as single-hop and multi-hop document question answering, pinpoint retrieval, variable tracking, and tool use. The method can be adapted into an efficient sparse variant, achieving near-constant encoding costs while maintaining (or sometimes improving) performance.
AuthorsWilliam L. Tong†**, Aryo Lotfi, Emmanuel Abbe, Josh Susskind, Cengiz Pehlevan†, Eran Malach
This paper was accepted at the Workshop on Personalized, Aligned, Long-term Memory for AI Systems (PALM) at NeurIPS 2026.
State space models (SSMs) combine fast inference and fixed state size with competitive language-modeling performance. Recently, state souping has been proposed as a procedure whereby an SSM independently encodes chunks of tokens like documents, tool definitions, or memory records. The resulting state encodings may be cached and recombined to cheaply construct a context for downstream queries. However, both state souping and vanilla SSMs deteriorate on long, crowded contexts, limiting their utility for long-horizon domains. To address this issue, we propose weighted state souping, a simple method that learns query-dependent weights for composing cached states. The resulting models improve long-context performance substantially across both single- and multi-hop document question answering, needle-in-a-haystack retrieval, variable tracking, and tool use. Our method can be further adapted into an efficient sparse variant, yielding near-constant encoding cost with respect to context length while preserving (and sometimes improving) performance. We also present simple theoretical arguments that illustrate why souping works and when it may fail, and discuss informally the advantages of weighted souping. Weighted state souping thus combines reusable context representations with selective retrieval, offering an efficient approach to language modeling over large collections of documents, tools, and other forms of structured context that increasingly dominate modern agentic workloads.
- † Harvard University
- ** Work done while at Apple
Related readings and updates.
Long-context handling remains a core challenge for language models: even with extended context windows, models often fail to reliably extract, reason over, and use the information across long contexts. Recent works like Recursive Language Models (RLMs) have approached this challenge by agentic way of decomposing long contexts into recursive sub-queries through programmatic interaction at inference. While promising, the success of RLMs critically…
State Space Models (SSMs) have become the leading alternative to Transformers for sequence modeling. Their primary advantage is efficiency in long-context and long-form generation, enabled by fixed-size memory and linear scaling of computational complexity. We begin this work by showing a simple theoretical result stating that SSMs cannot accurately solve any “truly long-form” generation problem (in a sense we formally define), undermining…
Source: Apple Machine Learning Research · machinelearning.apple.com