mlsanityData to 14 Oct 2026

Empathy Is Not the Scapegoat for Declines in Factuality and Reliability

arXiv 2610.18541 ↗First posted 13 Oct 2026cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

Language models are increasingly expected to support users in stressful or uncertain situations, where warmth and empathy are desirable. While recent work associates warmer models with reduced reliability, we show that this conclusion is incomplete. Prior studies often train primarily on empathetic dialogue, making it difficult to separate tone effects from SFT-induced distribution shift. We revisit this issue at both the model and evaluation levels. We study prompting, SFT, and RL for inducing warm behavior, and comprehensively evaluate capability preservation using an 11-dataset benchmark suite and EMOFACT under emotional contexts. Across multiple model families, we find that warmth effects are task- and model-dependent, while task-aligned adaptive tone can improve empathy and targeted reliability, although some capability trade-offs remain. We further introduce human-validated empathy evaluators specifically developed to assess empathetic model responses. Our results suggest that models can be more supportive in social interactions while remaining reliable.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 23 of its 40 references that are in the map. First posted on arXiv 13 Oct 2026.

Builds on

23 of 40 references.

References · 23

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  4. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,626
  5. 2021Evaluating Large Language Models Trained on Codecited by 8,320
  6. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  7. 2024Qwen2.5 Technical Reportcited by 4,942
  8. 2025Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilitiescited by 3,796
  9. 2021TruthfulQA: Measuring How Models Mimic Human Falsehoodscited by 2,975
  10. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,955
  11. 2020What Disease does this Patient Have? A Large-scale Open Domain Question Answering Dataset from Medical Examscited by 1,356
  12. 2021DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharingcited by 1,254
11 more
  1. 2023Steering Llama 2 via Contrastive Activation Additioncited by 1,085
  2. 2023Towards Understanding Sycophancy in Language Modelscited by 1,040
  3. 2023BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference Datasetcited by 860
  4. 2022Discovering Language Model Behaviors with Model-Written Evaluationscited by 836
  5. 2018Personalizing Dialogue Agents: I have a dog, do you have pets too?cited by 828
  6. 2021Towards Emotional Support Dialog Systemscited by 299
  7. 2024PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preferencecited by 261
  8. 2025The MASK Benchmark: Disentangling Honesty From Accuracy in AI Systemscited by 72
  9. 2025CounselBench: A Large-Scale Expert Evaluation and Adversarial Benchmarking of Large Language Models in Mental Health Question Answeringcited by 21
  10. 2025HumT DumT: Measuring and controlling human-like language in LLMscited by 12
  11. 2025Beacon: Single-Turn Diagnosis and Mitigation of Latent Sycophancy in Large Language Modelscited by 3

Built on by

Not cited yet.