Machine interpretable documents cut search and reading costs drastically
Machine-Interpretable Information: Compiling Documents into Searchable and Readable Protocol States
Computation and LanguageMachine Learning
Summary
Long documents are hard for language models to search and understand because reading lots of words uses a lot of computer power. The authors created a way to turn documents into tiny, fixed-size summaries called Machine-Interpretable Information (MII) that different AI systems can share and read quickly. This method reduces the effort needed to search and reason over documents, while still letting machines recreate important details when needed. Their approach works across different AI models and keeps important facts easy to find and reuse without reading everything again.
What this means in practice
- •For chatbot developers: Reduce the processing cost for chatbots by compiling knowledge documents into small, reusable memory states for quick retrieval and reasoning.
- •For enterprise search teams: Enable efficient and precise search in large document collections by using a shared lightweight document state format compatible with multiple AI models.
Authors
Yifan Wang, Dejing Dou
Abstract
Long-context language models interface with external knowledge through raw natural language. In retrieval-augmented systems, this creates a persistent index-payload schism: dense vectors enable searchable routing, but models must re-ingest lengthy text payloads for reasoning at O(N^2) attention cost. Existing compression methods further produce private states tied to specific architectures. We introduce Machine-Interpretable Information (MII), the first agent-to-agent (A2A) document-to-state protocol. A dual-timescale state-space Writer compiles documents into a canonical, fixed-bandwidth state (56 tokens), and a lightweight Translator maps it into any frozen Reader's embedding space, reducing query-time cost to O(K). The resulting .mii artifact unifies Retrieval (searchable geometry), Reasoning (global memory), and Reconstruction (grounded details) in a single transferable medium. We demonstrate strong cross-model interoperability across heterogeneous LLMs (e.g., Llama, Qwen, Mistral) -- despite the Writer using a legacy GPT-2 vocabulary, forcing genuine semantic translation rather than token-level memorization. Mechanistic probes reveal modular latent structure: entity representations can be causally traced and zero-shot transplanted between unrelated document states while remaining decodable. To address lexical reconstruction under fixed bandwidth, we propose Residual-MII, a cache hierarchy combining compiled global memory with sparse local evidence. On HotpotQA (7,405 queries), Residual-MII exceeds full-context Exact Match at approximately 7% of the attention FLOPs, suggesting a paradigm shift toward compiled, transferable neural document formats.