# Helpfulness Hurts: Domain-Dependent Degradation of Mid-Trained Compassion Values Under Post-Training

arXiv: https://arxiv.org/abs/2606.26102 · HTML: https://mlsanity.com/paper/2606.26102/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 16 of its 20 references that are in the map.
- First posted on arXiv 30 Apr 2026.

## Builds on (most cited first, 16 of 16)

- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!](https://mlsanity.com/paper/2310.03693/) (2023)
- [Editing Models with Task Arithmetic](https://mlsanity.com/paper/2212.04089/) (2022)
- [A General Language Assistant as a Laboratory for Alignment](https://mlsanity.com/paper/2112.00861/) (2021)
- [Towards Understanding Sycophancy in Language Models](https://mlsanity.com/paper/2310.13548/) (2023)
- [Discovering Language Model Behaviors with Model-Written Evaluations](https://mlsanity.com/paper/2212.09251/) (2022)
- [Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2307.15217/) (2023)
- [Safe RLHF: Safe Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2310.12773/) (2023)
- [Alignment faking in large language models](https://mlsanity.com/paper/2412.14093/) (2024)
- [Persona Vectors: Monitoring and Controlling Character Traits in Language Models](https://mlsanity.com/paper/2507.21509/) (2025)
- [Simple synthetic data reduces sycophancy in large language models](https://mlsanity.com/paper/2308.03958/) (2023)
- [Mitigating the Alignment Tax of RLHF](https://mlsanity.com/paper/2309.06256/) (2023)
- [Understanding Catastrophic Forgetting in Language Models via Implicit Inference](https://mlsanity.com/paper/2309.10105/) (2023)
- [Alignment Pretraining: AI Discourse Causes Self-Fulfilling (Mis)alignment](https://mlsanity.com/paper/2601.10160/) (2026)
- [Fundamental Safety-Capability Trade-offs in Fine-tuning Large Language Models](https://mlsanity.com/paper/2503.20807/) (2025)
- [Alignment midtraining for animals](https://mlsanity.com/paper/2604.13076/) (2026)

## Abstract (arXiv)

Standard post-training pipelines apply supervised fine-tuning (SFT) and reinforcement learning (RL) to make language models helpful, but these processes may inadvertently degrade values instilled during pre-training. We investigate whether the domain of post-training data differentially affects the retention of animal compassion values in a Llama 3.1 8B model mid-trained on compassion-oriented synthetic data, using both SFT (helpfulness via Dolly-15k vs. coding via Magicoder-110K) and GRPO (helpfulness via RLHFlow vs. coding via Magicoder), evaluated on the ANIMA 2.2 benchmark and MORU benchmark (Moral Reasoning Under Uncertainty). Helpfulness training significantly degrades animal compassion relative to coding training on ANIMA (SFT: 35.7% vs. 65.2%; GRPO: 18.7% vs. 32.0%), replicating across two independent helpfulness datasets and two training paradigms. On English MORU items, helpfulness training degrades general moral reasoning by 25.5 percentage points (46.4% vs. 71.9%), a striking gap that rivals the compassion effect in magnitude. However, this effect does not transfer cross-lingually: on the multilingual MORU benchmark, the domain effect disappears (SFT: 52.3% vs. 51.2%). In contrast, the animal compassion effect transfers consistently across languages, with Magicoder's ANIMA percentage-point gain over the base model 4.5 times larger on non-English items than English items. This divergence suggests that values instilled through mid-training are encoded more deeply and cross-lingually than reasoning improvements from domain-specific post-training. These results suggest that, for labs building on value-laden mid-training, coding-domain post-training may better preserve mid-trained values than helpfulness post-training without harming general reasoning capabilities.