When Cognitive Graphs Meet LLMs: BDEI Cognitive Pathways for Panic Emotional Arousal Prediction

2026-06-13 · Source: Computation and Language · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Robotics & Autonomous Systems, Data Science & Analytics · Depth: Expert, quick

Summary

PanicCognitivePath (PCP) is a novel framework designed to predict individual panic emotional arousal timing, addressing limitations in existing methods that fail to explicitly model the emotional arousal process. Grounded in appraisal emotion theory, PCP introduces a Psychological Safety Distance (PSD) model to unify multi-dimensional threat signals into a single risk metric. It integrates an explicit "Emotion" node into the traditional Belief-Desire-Intention (BDI) model, forming a Belief-Desire-Emotion-Intention (BDEI) pathway that directly couples threat appraisal with emotional arousal. This framework also strategically confines Large Language Models (LLMs) to parameter estimation for the Belief-to-Desire transition, thereby limiting hallucination risks and preventing error propagation. Experimental results on Hurricane Sandy data demonstrate that PCP improves arousal timing accuracy by 10.68% over baselines and reduces peak count error to 7.07%.

Key takeaway

For AI Scientists developing predictive models for human emotional states, this research suggests integrating explicit emotional pathways and structured cognitive models. You should consider confining large language models to specific, less critical parameter estimation tasks to mitigate hallucination risks. This approach, exemplified by PCP's BDEI pathway, can significantly improve arousal timing accuracy and reduce prediction errors in sensitive applications like emergency intervention.

Key insights

PCP integrates appraisal emotion theory, a PSD model, and a BDEI pathway with constrained LLMs to accurately predict panic arousal timing.

Principles

• Emotional arousal prediction benefits from explicit process modeling.
• Fusing multi-dimensional threat inputs yields unified risk perception.
• Confining LLM roles mitigates hallucination and error propagation.

Method

PCP uses a Psychological Safety Distance (PSD) model to map four-domain signals into a unified risk metric. Agents exceeding this threshold enter a Belief-Desire-Emotion-Intention (BDEI) pathway, where LLMs perform Belief-to-Desire parameter estimation.

In practice

• Apply PSD for unified risk assessment from diverse signals.
• Integrate explicit emotion nodes into cognitive models.
• Constrain LLMs to specific, less critical decision steps.

Topics

• Panic Prediction
• Emotional Arousal
• Cognitive Graphs
• Large Language Models
• BDEI Pathway
• Appraisal Emotion Theory

Best for: AI Scientist, Research Scientist, Machine Learning Engineer

Related on AIssential

• LLM-as-Judge Framework for Evaluating Tone-Induced Hallucination in Vision-Language Models — Linguistic tone systematically influences VLM hallucination, with varying effects on frequency and intensity across models and tasks.
• RECTOR: Masked Region-Channel-Temporal Modeling for Affective and Cognitive Representation Learning — RECTOR unifies region-channel-temporal EEG/sEEG representation learning via self-supervision, adapting functional regions for robust affective and cognitive disorder diagnosis.
• Emergence of Hierarchical Emotion Organization in Large Language Models — LLMs spontaneously organize emotions hierarchically, mirroring human psychology, but also replicate human biases in emotion recognition.
• Clinically Interpretable Sepsis Early Warning via LLM-Guided Simulation of Temporal Physiological Dynamics — LLM-guided temporal simulation improves sepsis early warning by providing interpretable physiological trajectory predictions.
• Interpreting and Enhancing Emotional Circuits in Large Vision-Language Models via Cross-Modal Information Flow — LVLMs process emotions through a "Adapt-Aggregate-Execute" mechanism with functionally decoupled sentiment-specific aggregation and emotion-general execution.
• Mental-R1: Aligning LLM Reasoning for Mental Health Assessment — Aligning LLM reasoning with human cognitive processes via specialized reinforcement learning improves mental health assessment accuracy.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Computation and Language.