Line of inquiry
Inquiring lines›How do training and design choices…›What makes certain prompting techn…›this line of inquiry
Why can't prompting alone inject genuinely new knowledge into models?
A broader line of inquiry — a family of 31 specific questions the research asks around this. Follow one into its inquiring-line page, or move sideways to a related line below.
Questions in this line of inquiry 31
Specific inquiring lines the field asks around this — ordered from the most general framing down to the most specific angle.
- Can prompt optimization alone inject knowledge models don't already have?
- Can prompting alone inject new domain knowledge into a model?
- Can prompt optimization inject genuinely new knowledge into a model?
- Can prompting inject new knowledge into already-trained AI models?
- Can users inject entirely new knowledge into models through prompting alone?
- What knowledge can prompt optimization actually activate in trained models?
- Can prompt optimization inject new knowledge into language models?
- Why does prompt optimization alone fail to inject genuinely new knowledge?
- Why does prompting discover capabilities that need reward-driven refinement?
- Can prompt engineering improve reasoning or only move requests into denser regions?
- Can prompt optimization or fine-tuning inject knowledge models do not already contain?
- How much knowledge can prompt optimization inject without retraining?
- How does explicit exploratory prompting compare to fine-tuned reinforcement learning for in-context adaptation?
- How does prompt context activation differ from parameter-based knowledge injection?
- Does joint optimization of prompts and parameters outperform separate tuning?
- Can steering internal features bypass or override prompt-level instructions in simulations?
- Can prompting unlock compositional skills that pretraining already learned?
- Can activation-space interventions reach biases that prompting cannot address?
- Can prompting-only specialization hide domain boundaries from users?
- Can runtime interventions like meta-cognitive prompting work where training interventions fail?
- How does prompt optimization differ from building persistent activation context?
- How do prompting and activation steering relate as compression strategies?
- Can activation steering directly steer models toward concise reasoning without prompting?
- Can activation-space directions reliably steer LLM behavior without retraining or prompting?
- Is prompt engineering a workaround rather than a capability fix?
- What happens when prompter skill matters more than domain expertise?
- What limits the capacity of context-based fast adaptation channels?
- What happens when prompt-optimized results lack anchoring in real data?
- How do slow weight updates and fast prompt updates interact in self-improvement?
- What makes a prompt update cheaper and more reversible than a weight update?
- How much does sliding-window augmentation improve single-session modeling?