Unveiling the Cognitive Compass: Theory-of-Mind–Guided Multimodal Emotion Reasoning

ICLR 2026 Conference Submission187 Authors

Keywords: Multimodal Affective Computing, Multimodal Understanding and Reasoning, Reinforcement Learning

TL;DR: We present HitEmotion, a ToM-grounded benchmark, and TMPO, a reinforcement learning approach with ToM-guided reasoning chains, to enhance emotional reasoning in multimodal LLMs.

Abstract: Despite rapid progress in multimodal large language models (MLLMs), their capability for deep emotional understanding remains limited. We argue that genuine affective intelligence requires explicit modeling of Theory of Mind (ToM), the cognitive substrate from which emotions arise. To this end, we introduce HitEmotion, a ToM-grounded hierarchical benchmark that diagnoses capability breakpoints across increasing levels of cognitive depth. Second, we propose a ToM-guided reasoning chain that tracks mental states and calibrates cross-modal evidence to achieve faithful emotional reasoning. We further introduce TMPO, a reinforcement learning method that uses intermediate mental states as process-level supervision to guide and strengthen model reasoning. Extensive experiments show that HitEmotion exposes deep emotional reasoning deficits in state-of-the-art models, especially on cognitively demanding tasks. In evaluation, the ToM-guided reasoning chain and TMPO improve end-task accuracy and yield more faithful, more coherent rationales. In conclusion, our work provides the research community with a practical toolkit for evaluating and enhancing the cognition-based emotional understanding capabilities of MLLMs.

Primary Area: datasets and benchmarks

Submission Number: 187