Emotion2Skill: Model-Internal Emotion Signals for Adaptive Skill Selection and Evolution
2026-08-10 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors looked at how language models (LLMs) pick skills to solve tasks and found that these decisions usually rely only on text cues, not the model's own feelings or 'emotions.' They created Emotion2Skill, a system that reads the internal emotional state of the LLM to help decide which skills to use and when to improve them. Testing showed this approach made the models better at completing tasks on two benchmarks compared to traditional methods. This work shows that internal emotion signals inside LLMs can guide better decision-making, not just explain behaviors after the fact.
Large Language Models (LLMs)Skill-based agentsInternal model stateEmotion representationSkill selectionRouting decisionsResidual streamEmotion2Skill frameworkBenchmark testingInterpretability
Authors
Bohan Lin, Hejia Geng, Xinyi Xie, Heng Zhou, Qinghua Xing, Bo Liu, Chen Zhang, Yudong Zhang
Abstract
Skill-based LLM agents select reusable procedures from an external library to solve complex tasks, yet their routing decisions rely entirely on text-level signals such as task descriptions, verbal reflections, and experience-derived rules, while the model's own internal representational state remains unobserved. Recent interpretability work has shown that LLMs maintain linear emotion representations that causally influence behavior; however, these representations have been exploited only for post-hoc analysis or direct output steering, and have not been used to inform agent-level decision-making. We propose Emotion2Skill, a framework that extracts LLM-internal emotion vectors and incorporates them into both skill selection and skill evolution. At each decision step, a 27-dimensional emotion state is extracted from the residual stream and mapped to a confidence-gated summary injected into the routing prompt. Beyond online selection, emotion trajectories are analyzed for abrupt internal-state shifts to pinpoint problematic skill invocations, guiding targeted SOP rewriting that replaces the coarse binary outcome signal of prior methods. On WebShop and ALFWorld, Emotion2Skill with Qwen3-8B improves over the Zero-Shot baseline by +26.9% success rate and +25.5% average success respectively, outperforming all baselines on both benchmarks with consistent gains on Qwen3-14B. Co-activation analysis further reveals semantically coherent emotion--skill pairings, confirming that the routing improvements reflect meaningful internal-state signals rather than opaque statistical correlations. These results establish LLM-internal emotion representations as an effective decision-level signal for orchestrating agent skill systems, extending their utility beyond interpretability and output steering. The code is available at https://github.com/BoHan-LIN04/Emotion2Skill.