RESEARCH · RESEARCH · #2540
DIVA: decision-informed information seeking for LLMs
A new arXiv paper (arXiv:2610.32181v1) introduces Decision-Informed Value-of-information Acquisition (DIVA), a training-free procedure that has an LLM specify a small numeric decision model and then values candidate checks by how much they change decision utility net of cost, stopping when no check is worth its cost; its variant DIVA-U additionally models uncertainty in those estimates. Across four domains and 20 LLMs, the authors report lower decision regret than all tested baselines (with a 75% regret reduction in synthetic scenarios where check reliabilities are known); code is available at github.com/seunghan96/diva.
KEY POINTS
- A new arXiv paper (arXiv:2610.32181v1) introduces Decision-Informed Value-of-information Acquisition (DIVA), a training-free procedure that has an LLM specify a small numeric decision model and then values candidate checks by how much they change decision utility net of cost, stopping when no check is worth its cost; its variant DIVA-U additionally models uncertainty in those estimates.
- Across four domains and 20 LLMs, the authors report lower decision regret than all tested baselines (with a 75% regret reduction in synthetic scenarios where check reliabilities are known); code is available at github.com/seunghan96/diva.
- This provides a simple, model-agnostic way to make LLMs stop unnecessary lookups and explicitly trade off information cost versus decision value, improving efficiency and decision quality in LLM-assisted tasks.
WHY IT MATTERS
This provides a simple, model-agnostic way to make LLMs stop unnecessary lookups and explicitly trade off information cost versus decision value, improving efficiency and decision quality in LLM-assisted tasks.