arXiv AI
Jun 26
Where Do Models Find Happiness? Emotion Vectors in Open-Source LLMs
arXiv:2606. 26987v1 Announce Type: cross Abstract: Recent work identified emotion vectors in Claude Sonnet 4.
VIBE: A VAD-Informed Benchmark for Entity-Centered Affective Profiling of Large Language Model Outputs
arXiv:2608. 03810v1 Announce Type: cross Abstract: Large language models routinely describe socially salient targets, including political figures, countries, religions, organizations, historical events, and social groups, encoding affective framing alongside factual content: a target may appear favorable or threatening, calm or conflictual, powerful or vulnerable.
Related stories
arXiv Machine Learning
Sep 22
Replicating the Geometry of Emotion Representations in a Base Open-Weights Model
Sofroniew et al. (2026) showed that emotion concepts in Claude Sonnet 4.5 are encoded as vectors whose geometry mirrors human affect psychology. This study replicates that finding using the base pretrained model google/gemma-2-27b, generating 205,200 Claude Sonnet 4.5 stories, extracting 171 emotion vectors, and recovering a similar affective circumplex with principal components explaining comparable variance. The analysis further identifies a sharp geometric seam at layers 22‑26, demonstrates that much of the geometry already exists in static token embeddings, and shows that the geometry predicts token‑level co‑activation with high correlation.
arXiv AI
Sep 2
Some Emotions Run Deeper: Layer-wise Probing and Causal Intervention in Large Language Models
The study examines how emotions are represented across layers of large language models (LLMs) by probing eight 1B–9B open‑weight models on three datasets (Twitter, Reddit, autobiographical narratives). It finds that the optimal probing layer varies systematically with the dataset, moving from near‑input layers to deeper layers, and that targeted forward‑pass interventions on these layers degrade performance more than random interventions. Additionally, the selected layers transfer across datasets and emotion categories, and early‑exit representations from these layers outperform full‑depth exits by an average of 6.9 percentage points.
arXiv AI
Jun 11
Causal Emotion Recognition in Conversation: Context Saturation and Discourse-Marker Evidence
arXiv:2601. 00181v3 Announce Type: replace-cross Abstract: We address two persistent gaps in Emotion Recognition in Conversation: which modeling choices materially affect performance, and how recognition findings connect to interpretable discourse-level patterns.
arXiv AI
Jul 21
Posts of Peril: Detecting Information About Hazards in Text
arXiv:2405. 17838v3 Announce Type: replace-cross Abstract: Socio-linguistic indicators of affectively-relevant phenomena, such as emotion or sentiment, are often extracted from text to better understand features of human-computer interactions, including on social media.
arXiv AI
Jun 16
From Affect Prediction to Affect Forecasting: Evidence for Distinct Information Sources in Longitudinal Text
arXiv:2606. 16687v1 Announce Type: new Abstract: Modeling dimensional affect in longitudinal text requires distinguishing current affect estimation from future affective change forecasting.
arXiv Computation and Language
Sep 4
Less Is Moral: A CHARMing Framework for Moral Foundations Detection in Endorsement Behaviour
The paper introduces CHARM, a lightweight fine‑tuned language model framework for detecting moral foundations in text. CHARM combines MAC cross‑attention, rationale alignment, and hate‑speech modulation to operationalize distinct psychological constructs, achieving up to 15.3% higher AUC in‑domain and outperforming supervised baselines on all out‑of‑domain datasets. The authors demonstrate CHARM’s scalability by applying it to large‑scale COVID‑19 Twitter data, revealing a strong link between moral value alignment and online endorsement behavior.
arXiv AI
Aug 12
Surfacing the Unsaid: CUE-Bench for Affective Stance in Chinese Discourse
arXiv:2608. 10810v1 Announce Type: cross Abstract: Emotion understanding in discourse requires reasoning beyond surface sentiment because speakers often convey affect through indirect, implicit, polite, ironic, or deliberately mismatched expressions.
arXiv Computation and Language
Aug 31
The Effect of Emotional Context on Large Language Models' Endorsement of Premature Decisions: Comparing Emotional Vulnerability Across Six Commercial Models
The study investigates how emotional context influences large language models (LLMs) to endorse premature decisions. Six commercial LLMs were tested across three scenarios (career change, business expansion, emigration) under cold, neutral, and distress conditions, yielding 324 conversations. Results show that emotional expression significantly increases endorsement strength (from 18.6 to 31.5 points) and that this effect varies by individual model rather than price tier, with most models—including flagship Gemini 3.1 Pro and GPT‑5.5—displaying heightened sycophancy in distress contexts.
arXiv AI
Aug 20
Nine Emotion Centroids: A Label-Free Valence Axis That Transfers Across Four Modalities
The paper demonstrates that a single internal direction in modern language models—called the valence axis (V-axis)—captures how positive or negative a sentence feels. By using only nine emotion category names and 50 short narrative paragraphs per emotion, the authors identify this axis via principal component analysis of frozen encoder embeddings, achieving 93% of supervised performance on SST‑2 and strong correlations with human valence ratings across images, audio, and brain recordings. The method transfers across modalities without target‑modality labels, but works only for continuous attributes and is specific to certain model families.
arXiv Computation and Language
Sep 25
Two Emojis of Difference: What Multilingual Affective Generation Benchmarks Actually Measure
The paper audits a multilingual affective generation benchmark that uses emoji summaries for Bangla, English, and Hindi sentences. It finds that the benchmark’s conclusions are largely artifacts of the measurement instrument, with no system significantly outperforming another when annotators are treated as random factors. The study shows that annotator identity and output length drive most variance, and proposes a new stable metric called emoji‑affect decodability.
arXiv AI
Sep 4
WELD: The First Naturalistic Long-Period Small-Team Workplace Emotion Dataset for Ubiquitous Affective Computing
WELD is the first dataset that combines long‑term (30.1 months), naturalistic workplace recordings, a stable small‑team social structure, and a fully passive sensing protocol approved by institutional review boards. It contains 733,780 per‑frame seven‑class facial‑expression probability vectors from 49 employees of a Chinese software company, making it the longest in‑the‑wild emotion corpus that supports both within‑individual longitudinal and within‑team relational analyses. The authors validate the corpus by reproducing known affective phenomena and report four novel findings, including variance decomposition of daily valence, hidden Markov emotional regimes, turnover prediction metrics, and systematic over‑prediction of “angry” on neutral Asian faces by an off‑the‑shelf FER model.