RECAST: Learning to Compute the Right Context through Adaptive Evidence Routing
Novel approach to evidence construction in RAG systems with demonstrated performance improvements over baseline models.
AI Summary
Researchers introduce RECAST, a learned framework that formulates evidence construction as sequential decision process over retrieval and computation operations.
Excerpt
Large language models are increasingly applied to tasks grounded in long, heterogeneous information sources. Conventional Retrieval-Augmented Generation (RAG) relies on fixed similarity-based retrieval, while agentic variants adapt queries and tool use but remain largely retrieval-centric. However, in many tasks, the evidence required for a solution is not explicitly present in any single source item. Instead, it must be derived through filtering, aggregation, or computation across multiple sour