Publikation
Learning Explicit Behavioral Models with Adaptive Questions and World-Model Probes
Hikaru Shindo; Yu Deng; Teng Cao; Quentin Delfosse; Christopher Tauchmann; Jannis Blüml; Gopika Sudhakaran; Kristian Kersting
In: Computing Research Repository eprint Journal (CoRR), Vol. abs/2606.07127, Pages 1-26, arXiv, 2026.
Zusammenfassung
Interactive agents trained only against task return can achieve high scores while
failing to represent the mechanisms that make their actions succeed. This makes
brittle behavior difficult to diagnose and limits adaptation when environment
dynamics change. Existing LLM reflection and policy-code repair can revise
behavior from failed trajectories, but questions and world-understanding tests
are usually used only after training. We introduce an Explicit Symbolic Behav-
ioral Model (ESBM), a trainable behavioral model that couples task performance
with evidence-grounded question answering and executable mechanism predic-
tion. An ESBM represents behavior through typed predicates, weighted rules,
bounded options and mechanism memory; the mechanism layer predicts symbolic
events, object changes, rewards and terminal consequences under action inter-
ventions. After each rollout, adaptive questions and active world-model probes
convert score failures, QA errors and transition-prediction errors into constraints
for local ESBM edits. Candidate models are selected by a multi-criterion rule
that jointly evaluates task score, answerability and active world-model consis-
tency. Under the tested Atari-style protocols, ESBM learns high-scoring policies
while producing explicit answers and executable mechanism predictions, indi-
cating that adaptive questions can serve as both training pressure and reusable
benchmarks for mechanistic policy learning in this setting.