Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note2 min read

Can Induced Emotion Bias LLM Behaviors in Sequential Decision Making?

Source

arXiv

arxiv.org/abs/2607.12631v1

Read the source

Source

Plain-Language Brief

Researchers explored whether inducing emotions in large language models (LLMs) can influence their sequential decision-making using the Iowa Gambling Task (IGT). They found that while LLMs could sense emotions and learn from interactions, anger induction caused less sensitivity to penalties for bad decisions and reduced exploration early in the game. However, on average, induced emotion did not significantly bias the decision dynamics of LLM agents compared to human behavior.

What the Document Says

  • The study aimed to understand if induced emotions can affect the sequential decision-making of Large Language Models (LLMs).
  • The Iowa Gambling Task (IGT) was used, combined with an emotion induction procedure based on imagination.
  • It was confirmed that LLMs could perceive and resonate with strong, distinguishable emotions from context, and learn from sequential interactions at a human-like pace.
  • No significant bias in decision dynamics of LLM agents due to induced emotion on average was found.
  • Anger induction led to less sensitivity to penalties for bad decisions and reduced exploration early in the game for LLM agents.
  • These findings suggest subtle differences in emotional influence on LLM decision-making compared to human behavior.

Key Concepts Mentioned

  • Large Language Models (LLMs)
  • Iowa Gambling Task (IGT)
  • emotion induction
  • anger
  • decision-making under uncertainty

Claims Made by the Document

  • LLMs can perceive and resonate with emotions.
  • LLMs can learn from sequential interactions at a human-like pace.
  • Induced emotion does not significantly bias the decision dynamics of LLM agents on average.

Evidence Used

  • confirming that LLMs can sense strong, distinguishable emotions and learn from sequential interactions

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

As Large Language Models (LLMs) are increasingly deployed as autonomous agents in high-stakes domains, understanding contextual factors that may modulate their decision-making becomes critical. While LLMs are trained to perceive and resonate with users' emotions, it remains unclear whether induced emotion can influence their sequential decision-making. We investigate this question using the Iowa Gambling Task (IGT), a classic psychological paradigm for studying decision-making under uncertainty, combined with an imagination-based emotion induction procedure. We first validate the feasibility of this paradigm by confirming that LLMs can sense strong, distinguishable emotions from context and that LLM agents can learn from sequential interactions in a human-like pace. With the validated setup, we find that, different from humans, induced emotion does not significantly bias the decision dynamics of LLM agents on average. However, the effects of anger are conditioned: inducing anger makes LLM agents less sensitive to penalties for bad decisions, and in early stages of the game, anger can lower exploration, locking decisions into a few choices early. These findings reveal the subtle yet distinct effects of induced emotion on LLM decision-making compared to human behavior, and provide a tool for future research on affective modulation of LLM agents.