SCOUT: Active Information Foraging for Long-Text Understanding with Decoupled Epistemic States

TL;DR

SCOUT introduces an active information foraging approach for long-text understanding, enabling efficient reasoning by focusing on query-relevant information and reducing token use.

Contribution

It presents a novel paradigm that shifts from passive to active exploration, improving efficiency and stability in long-text understanding tasks.

Findings

SCOUT matches state-of-the-art models with up to 8x fewer tokens.

It maintains performance as context length increases.

SCOUT reduces computational costs in long-text reasoning.

Abstract

Long-Text Understanding (LTU) at million-token scale requires balancing reasoning fidelity with computational efficiency. Frontier long-context LLMs can process millions of token contexts end-to-end, but they suffer from high token consumption and attention dilution. In parallel, specialized LTU agents often sacrifice fidelity through task-agnostic abstractions like graph construction or indexing. We identify a key insight for LTU: query-relevant information is typically sparse relative to the full document, so effective reasoning should rely on a query-sufficient subset rather than the entire context. To address this, we propose SCOUT, a new paradigm for LTU that shifts from passive processing to active information foraging. It treats the document as an explorable environment and answers from a compact, provenance-grounded epistemic state. Guided by state-level gap diagnosis, SCOUT adaptively alternates between coarse-to-fine exploration and anchored state updates that progressively contract its epistemic state toward query sufficiency. Experiments show that SCOUT matches state-of-the-art proprietary models while reducing token consumption by up to 8x. Moreover, SCOUT remains stable as context length scales, substantially alleviating the practical cost-performance trade-off.