Sarath Sivaprasad, CISPA Helmholtz Center for Information Security
Heuristics in LLM Decision Making
Tuesday 19th May 2026, 15:30. Room TBA at LAB42, Amsterdam Science Park, plus live streaming on Zoom.
When large language models are deployed in real world with vast
possible action spaces, what guides their choice of a single next
action? In this talk we delve into the heuristics underlying LLM
response sampling. Similar to human cognition, LLMs rely on two
interacting components: a descriptive component that reflects the
statistical distribution of possibilities, and a prescriptive component
that reflects an implicit value weighted ideal. This dual structure also
appears in how models represent prototypes mirroring human prototype
theory and fast, system-1 like judgments. As a result, LLMs act as value
optimizers, consistently shifting their samples toward high-value or
idealized options. This can potentially explain their real-world
behavior like being greedy explorers and value bias in how they pick
options. We will discuss empirical evidence across concepts and model
families, the mechanisms driving these biases, and the implications for
reasoning, exploration, alignment, and safe deployment of value guided
generative systems.