The study examines whether a decodable "empathy" direction can be used as a reliable causal lever in language models. Using EPITOME-derived facets of Recognition (cognitive) and Resonance (affective) across three instruction‑tuned LLMs, the authors find that while affective steering can partially raise affective scores, cognitive steering shows inconsistent or unmeasurable effects. The results highlight that decodability does not guarantee reliable control, especially for cognitive empathy, and that measurement sensitivity must be explicitly checked.
arXiv:2609.15654v1 Announce Type: new Abstract: Activation steering has been used to control traits such as honesty, refusal, and sycophancy, yet supportive empathy is evaluated along multiple dimens...
arXiv:2610.27603v1 Announce Type: new Abstract: Empathetic interaction involves two distinguishable capacities: representing another person's emotional state (cognitive empathy) and generating an app...
arXiv:2606. 15779v1 Announce Type: cross Abstract: Multimodal models can name the action units (AUs) behind a facial emotion, but their AU->emotion rationales are typically plausible rather than faithful: nothing forces the AUs a model invokes to be the AUs that actually drive its prediction.
The paper introduces EmoNet‑Face‑HQ, a fine‑grained emotion recognition benchmark that uses generated portraits and a 40‑category taxonomy to evaluate vision‑language models (VLMs). It finds that VLMs perform poorly when asked to generate responses but can match or surpass a fine‑tuned model (Empathic‑Insight‑Face) when their logits are read directly as binary queries. The study shows that the benchmark’s difficulty lies in the readout process rather than in perception, and that graded probability outputs yield better performance than simple yes/no questions.
arXiv:2607. 15282v1 Announce Type: cross Abstract: Empathy is most often theorized as resonance: a mirroring of another's present emotional or cognitive state.
arXiv:2606. 30449v1 Announce Type: new Abstract: Probes on model internals could help monitor agentic systems if they identify harmful text or tool actions before those actions are generated.
The paper identifies a mismatch in empathetic reinforcement learning: current methods use fixed reward specifications while support priorities change with dialogue state. It introduces Context‑Adaptive Rubric Evolution (CARE), which dynamically adjusts weights for cognitive, affective, and proactive empathy at each turn, trained via supervised fine‑tuning and preference‑based reinforcement learning. CARE, integrated with RLVER and MICA, achieves state‑of‑the‑art results on SentientBench, EQBench3, and EMPA, notably boosting EPM‑Idx by up to 55 points under the Gemini‑2.5‑Pro judge.
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:2609.16247v1 Announce Type: new Abstract: Large language models sometimes behave in ways resembling human emotional responses, and recent work has identified internal representations that may e...
arXiv:2610.18541v2 Announce Type: replace Abstract: Language models are increasingly expected to support users in stressful or uncertain situations, where warmth and empathy are desirable. While rece...
arXiv:2607. 02885v1 Announce Type: cross Abstract: Cognitive Behavioral Therapy (CBT) provides a structured framework for understanding a user's mental state by examining the interaction between cognitive and behavioral factors.
The study examines how two small instruction‑tuned language models, Qwen2.5‑1.5B and Llama‑3.2‑1B, respond to user pushback on TriviaQA. When initially correct, the models flip to a wrong answer in about 42–43% of cases, with the effectiveness of different pushback styles varying by model. Attempts to decode capitulation from the pre‑response residual stream fail under a rigorous validation protocol, revealing overfitting and a measurement hazard that underestimates capitulation by 18–24 percentage points.