Fine-tuning enables a source model to acquire desired capabilities and behaviors in a target domain while retaining much of its general-purpose competence. However, this adaptation process can also degrade alignment properties that were present in the source model.
arXiv:2608. 04347v1 Announce Type: new Abstract: Fine-tuning enables a source model to acquire desired capabilities and behaviors in a target domain while retaining much of its general-purpose competence.
Warning: This paper studies stereotypes and biases, and contains potentially disturbing examples, used for illustration purposes only. Our findings should not be interpreted as an argument against alignment.
arXiv:2607. 22676v1 Announce Type: new Abstract: Post-training is a key mechanism for adapting large language models to downstream tasks.
arXiv:2606. 19168v1 Announce Type: new Abstract: To achieve deeper safety alignment for large language models (LLMs), recent efforts have studied how to push safety interventions earlier into the pretraining stage, primarily by filtering unsafe data or rewriting it into safer forms.
arXiv:2511. 00382v2 Announce Type: replace-cross Abstract: Organizations increasingly adapt Large Language Models (LLMs) from public repositories such as HuggingFace to downstream tasks.
arXiv:2606. 31591v1 Announce Type: cross Abstract: Emergent misalignment (EM) is a recently discovered phenomenon in LLMs where fine-tuning on a narrow misaligned task, such as writing insecure code, leads to broadly misaligned behaviour on unrelated prompts.
Emergent misalignment (EM) is the phenomenon where fine-tuning a language model on a narrow task leads to harmful behavior in unrelated domains. A leading mechanistic account attributes EM to persona features: latent directions acquired during pre-training that misaligned fine-tuning amplifies.
Modern LLMs are alarmingly susceptible to surprisingly simple immaterial changes of input prompts: a casual hint, an incorrectly labeled few-shot example, or a fake prior assistant turn often flips an originally correct answer. We study where this susceptibility, spanning sycophancy and related cue-induced biases, lives inside the model.
arXiv:2606. 08682v1 Announce Type: cross Abstract: Activation steering has emerged as a popular inference-time technique for modulating the behavior of large language models (LLMs).
arXiv:2607. 26173v1 Announce Type: new Abstract: Alignment training, model organisms, and toy models are usually treated as separate research areas.
arXiv:2607. 18114v1 Announce Type: cross Abstract: Modern LLMs are alarmingly susceptible to surprisingly simple immaterial changes of input prompts: a casual hint, an incorrectly labeled few-shot example, or a fake prior assistant turn often flips an originally correct answer.
arXiv:2608. 13069v1 Announce Type: new Abstract: Large language models (LLMs) are predominantly aligned to function as passive, sycophantic assistants.