Proposes a decision-centric architecture that separates signal estimation from control policy to make LLM system decisions explicit and inspectable.
April 2, 2026
Original Paper
Decision-Centric Design for LLM Systems
arXiv · 2604.00414
The Takeaway
By disentangling 'whether to act' from 'what to generate,' this framework allows developers to diagnose whether a failure was due to bad information (signal) or bad strategy (policy), leading to significantly more reliable agentic systems.
From the abstract
LLM systems must make control decisions in addition to generating outputs: whether to answer, clarify, retrieve, call tools, repair, or escalate. In many current architectures, these decisions remain implicit within generation, entangling assessment and action in a single model call and making failures hard to inspect, constrain, or repair. We propose a decision-centric framework that separates decision-relevant signals from the policy that maps them to actions, turning control into an explicit