reference-world-models

Tag

Cards List
#reference-world-models

HalluWorld: A Controlled Benchmark for Hallucination via Reference World Models

arXiv cs.CL · 2026-05-20 Cached

HalluWorld is a controlled benchmark framework for evaluating hallucination in large language models using explicit reference world models across synthetic environments like gridworlds, chess, and realistic terminal tasks. It enables fine-grained analysis of failure modes such as perceptual hallucination, multi-step state tracking, and causal simulation, revealing that frontier models still struggle with complex reasoning not solved by extended thinking.

0 favorites 0 likes
← Back to home

Submit Feedback