Thought, the LLM can verbalize that the unknown city is Paris and use this fact to answer downstream questions.

The in-context learning (ICL) capability of pre-trained models based on the transformer architecture has received growing interest in recent years.

Recent language models, such as GPT -3+ [Brown et al., 2020, Achiam et al., 2023], have demonstrated Recent attempts to understand in-context learning have focused on various aspects. On the practical side, research has investigated the impact of different factors on in-context learning.

Longer contexts present a new attack surface for adversarial attacks. In search of a "fruit-fly" of long-context vulnerabilities, we study Many-shot Jailbreaking (MSJ; Figure 1), a simple yet effective and scalable jailbreak.

Each environment state has a corresponding array and a scene graph representation.

Large language models (LLMs) have demonstrated remarkable capabilities in understanding and generating natural language data.