Lune

ACL Findings2025

Mechanistic Interpretability of Emotion Inference in Large Language Models

Ala N. Tak, Amin Banayeeanzade, Anahita Bolourani, Mina Kian, Robin Jia, Jonathan Gratch

Abstract

Large language models (LLMs) show promising capabilities in predicting human emotions from text.However, the mechanisms through which these models process emotional stimuli remain largely unexplored.Our study addresses this gap by investigating how autoregressive LLMs infer emotions, showing that emotion representations are functionally localized to specific regions in the model.Our evaluation includes diverse model families and sizes, and is supported by robustness checks.We then show that the identified representations are psychologically plausible by drawing on cognitive appraisal theory-a well-established psychological framework positing that emotions emerge from evaluations (appraisals) of environmental stimuli.By causally intervening on construed appraisal concepts, we steer the generation and show that the outputs align with theoretical and intuitive expectations.This work highlights a novel way to causally intervene and control emotion inference, potentially benefiting safety and alignment in sensitive affective domains.Code at: GitHub repo.# Prompt Template 1 What are the inferred emotions in the following contexts?Context: My first child was born.Answer: joy Context: My dog died last week.Answer: sadness

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

5 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • How Controllable Are Large Language Models? A Unified Evaluation across Behavioral Granularities cites Inference-Time Intervention: Eliciting Truthful Answers from a Language Model
  • How does GPT-2 compute greater-than?: Interpreting mathematical abilities in a pre-trained language model cites Investigating Gender Bias in Language Models Using Causal Mediation Analysis
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers5

  1. Good Arguments Against the People Pleasers: How Reasoning Mitigates (Yet Masks) LLM SycophancyZhaoxin Feng, Zheng Chen, Jianfei Ma, Yip Tin Po et al.ACL 2026 · 3 citations
  2. Interpreting and Enhancing Emotional Circuits in Large Vision-Language Models via Cross-Modal Information FlowChengsheng Zhang, Chenghao Sun, Zhining Xie, Xinmei TianICML 2026
  3. Reconsidering LLM Uncertainty Estimation Methods in the WildYavuz Faruk Bakman, Duygu Nur Yaldiz, Sungmin Kang, Tuo Zhang et al.ACL 2025
  4. How Controllable Are Large Language Models? A Unified Evaluation across Behavioral GranularitiesZiwen Xu, Kewei Xu, Haoming Xu, Haiwen Hong et al.ACL 2026
  5. Are Emotion and Rhetoric Neurons in LLM? Neuron Recognition and Adaptive Masking for Emotion-Rhetoric Prediction SteeringLi Zheng, Xin Zhang, Shuyi He, Fei Li et al.ACL 2026

Builds on17

  1. Inference-Time Intervention: Eliciting Truthful Answers from a Language ModelKenneth Li, Oam Patel, Fernanda B. Viégas, Hanspeter Pfister et al.NeurIPS 2023 · 1,549 citations
  2. Towards Automated Circuit Discovery for Mechanistic InterpretabilityArthur Conmy, Augustine N. Mavor-Parker, Aengus Lynch, Stefan Heimersheim et al.NeurIPS 2023 · 861 citations
  3. Investigating Gender Bias in Language Models Using Causal Mediation AnalysisJesse Vig, Sebastian Gehrmann, Yonatan Belinkov, Sharon Qian et al.NeurIPS 2020 · 851 citations
  4. The Linear Representation Hypothesis and the Geometry of Large Language ModelsKiho Park, Yo Joong Choe, Victor VeitchICML 2024 · 461 citations
  5. How does GPT-2 compute greater-than?: Interpreting mathematical abilities in a pre-trained language modelMichael Hanna, Ollie Liu, Alexandre VariengienNeurIPS 2023 · 251 citations

Related papers

  1. Beyond Context to Cognitive Appraisal: Emotion Reasoning as a Theory of Mind Benchmark for Large Language ModelsGerard Christopher Yeo, Kokil JaidkaACL Findings 2025 · 3 citations
  2. Emotions Where Art Thou: Understanding and Characterizing the Emotional Latent Space of Large Language ModelsBenjamin Z. Reichman, Adar Avsian, Larry HeckICLR 2026 · 14 citations
  3. Do LLMs “Feel”? Emotion Circuits Discovery and ControlChenxi Wang, Yixuan Zhang, Ruiji Yu, Yufei Zheng et al.ICML 2026 · 11 citations
  4. Apathetic or Empathetic? Evaluating LLMs' Emotional Alignments with HumansJen-tse Huang, Man Ho Lam, Eric John Li, Shujie Ren et al.NeurIPS 2024 · 63 citations
  5. Uncovering Sentiment Analysis Circuit in Large Language ModelShichen Li, Zhouyang Wang, Zhongqing Wang, Peifeng LiACL 2026
Dusk over the sea between two cliffs drawn in fine vertical lines