How Emotion Shapes the Behavior of LLMs and Agents: A Mechanistic Study

• URL: http://arxiv.org/abs/2604.00005v1
• Date: Mon, 09 Mar 2026 12:20:02 GMT
• Title: How Emotion Shapes the Behavior of LLMs and Agents: A Mechanistic Study
• Abstract summary: We investigate whether emotional signals can shape the behavior of large language models (LLMs) and agents.<n>Existing emotion-aware studies mainly treat emotion as a surface-level style factor or a perception target.<n>We propose E-STEER, an interpretable emotion steering framework that enables direct representation-level intervention.
• Score: 33.5007842598689
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Emotion plays an important role in human cognition and performance. Motivated by this, we investigate whether analogous emotional signals can shape the behavior of large language models (LLMs) and agents. Existing emotion-aware studies mainly treat emotion as a surface-level style factor or a perception target, overlooking its mechanistic role in task processing. To address this limitation, we propose E-STEER, an interpretable emotion steering framework that enables direct representation-level intervention in LLMs and agents. It embeds emotion as a structured, controllable variable in hidden states, and with it, we examine the impact of emotion on objective reasoning, subjective generation, safety, and multi-step agent behaviors. The results reveal non-monotonic emotion-behavior relations consistent with established psychological theories, and show that specific emotions not only enhance LLM capability but also improve safety, and systematically shape multi-step agent behaviors.

Related papers

• From Syntax to Emotion: A Mechanistic Analysis of Emotion Inference in LLMs [1.4692712025142225]
  We investigate the internal mechanisms of emotion recognition in large language models (LLMs)<n>By analyzing sparse feature activations across layers, we identify a consistent three-phase information flow.<n>We show that emotion representations comprise both shared features across emotions and emotion-specific features.
  arXiv  Detail & Related papers  (2026-04-28T17:03:16Z)
• EMO-R3: Reflective Reinforcement Learning for Emotional Reasoning in Multimodal Large Language Models [62.3977734456669]
  We propose Reflective Reinforcement Learning for Emotional Reasoning (EMO-R3), a framework designed to enhance the emotional reasoning ability of Multimodal Large Language Models (MLLMs)<n>We introduce Structured Emotional Thinking to guide the model to perform step-by-step emotional reasoning in a structured and interpretable manner, and design a Reflective Emotional Reward that enables the model to re-evaluate its reasoning based on visual-text consistency and emotional coherence.<n>EMO-R3 significantly improves both the interpretability and emotional intelligence of MLLMs, achieving superior performance across multiple visual emotional understanding benchmarks.
  arXiv  Detail & Related papers  (2026-02-27T08:42:52Z)
• Do LLMs "Feel"? Emotion Circuits Discovery and Control [54.57583855608979]
  We study the internal mechanisms that give rise to emotional expression and in controlling emotions in generated text.<n>This is the first systematic study to uncover and validate emotion circuits in large language models.
  arXiv  Detail & Related papers  (2025-10-13T12:24:24Z)
• Beyond Emotion Recognition: A Multi-Turn Multimodal Emotion Understanding and Reasoning Benchmark [15.900703216919169]
  We introduce a multi-turn multimodal emotion understanding and reasoning benchmark, which encompasses 1,451 video data from real-life scenarios, along with 5,101 progressive questions.<n>We propose a multi-agent framework, where each agent specializes in a specific aspect, such as background context, character dynamics, and event details, to improve the system's reasoning capabilities.
  arXiv  Detail & Related papers  (2025-08-23T01:10:29Z)
• AI with Emotions: Exploring Emotional Expressions in Large Language Models [0.0]
  Large Language Models (LLMs) play role-play as agents answering questions with specified emotional states.<n>Russell's Circumplex model characterizes emotions along the sleepy-activated (arousal) and pleasure-displeasure (valence) axes.<n> evaluation showed that the emotional states of the generated answers were consistent with the specifications.
  arXiv  Detail & Related papers  (2025-04-20T18:49:25Z)
• Mechanistic Interpretability of Emotion Inference in Large Language Models [16.42503362001602]
  We show that emotion representations are functionally localized to specific regions in large language models.<n>We draw on cognitive appraisal theory to show that emotions emerge from evaluations of environmental stimuli.<n>This work highlights a novel way to causally intervene and precisely shape emotional text generation.
  arXiv  Detail & Related papers  (2025-02-08T08:11:37Z)
• From Rational Answers to Emotional Resonance: The Role of Controllable Emotion Generation in Language Models [16.350658746140788]
  Large language models (LLMs) struggle to express emotions in a consistent, controllable, and contextually appropriate manner.<n>We propose a controllable emotion generation framework based on Emotion Vectors (EVs)<n>Our method enables fine-grained, continuous modulation of emotional tone without any additional training or architectural modification.
  arXiv  Detail & Related papers  (2025-02-06T13:38:57Z)
• ECR-Chain: Advancing Generative Language Models to Better Emotion-Cause Reasoners through Reasoning Chains [61.50113532215864]
  Causal Emotion Entailment (CEE) aims to identify the causal utterances in a conversation that stimulate the emotions expressed in a target utterance. Current works in CEE mainly focus on modeling semantic and emotional interactions in conversations. We introduce a step-by-step reasoning method, Emotion-Cause Reasoning Chain (ECR-Chain), to infer the stimulus from the target emotional expressions in conversations.
  arXiv  Detail & Related papers  (2024-05-17T15:45:08Z)
• Enhancing Emotional Generation Capability of Large Language Models via Emotional Chain-of-Thought [50.13429055093534]
  Large Language Models (LLMs) have shown remarkable performance in various emotion recognition tasks. We propose the Emotional Chain-of-Thought (ECoT) to enhance the performance of LLMs on various emotional generation tasks.
  arXiv  Detail & Related papers  (2024-01-12T16:42:10Z)
• Large Language Models Understand and Can be Enhanced by Emotional Stimuli [53.53886609012119]
  We take the first step towards exploring the ability of Large Language Models to understand emotional stimuli. Our experiments show that LLMs have a grasp of emotional intelligence, and their performance can be improved with emotional prompts. Our human study results demonstrate that EmotionPrompt significantly boosts the performance of generative tasks.
  arXiv  Detail & Related papers  (2023-07-14T00:57:12Z)
• Emotion Recognition from Multiple Modalities: Fundamentals and Methodologies [106.62835060095532]
  We discuss several key aspects of multi-modal emotion recognition (MER) We begin with a brief introduction on widely used emotion representation models and affective modalities. We then summarize existing emotion annotation strategies and corresponding computational tasks. Finally, we outline several real-world applications and discuss some future directions.
  arXiv  Detail & Related papers  (2021-08-18T21:55:20Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.