Line of inquiry
Inquiring lines›How do design choices determine AI…›How do prompt design and interface…›this line of inquiry
Can prompting inject new knowledge into models or only activate existing capabilities?
A broader line of inquiry — a family of 64 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 64
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 new knowledge into language models?
- Can prompting inject new knowledge into already-trained AI models?
- Can users inject entirely new knowledge into models through prompting alone?
- Can prompt optimization inject genuinely new knowledge into a model?
- What knowledge can prompt optimization actually activate in trained models?
- Why does prompting discover capabilities that need reward-driven refinement?
- How does explicit exploratory prompting compare to fine-tuned reinforcement learning for in-context adaptation?
- Are instruction-tuned models more or less sensitive to prompt semantics than others?
- Does foundational model training or user priors more strongly shape final outputs?
- Why does prompt optimization alone fail to inject genuinely new knowledge?
- How do training-data priors influence model defaults when context is ambiguous?
- How does supervised finetuning change what prompting techniques actually accomplish?
- How do pretraining biases interact differently with prompts across model tiers?
- Why does instruction-tuning reduce a model's context-following behavior?
- Can input augmentation and rephrasing compensate for smaller model limitations?
- Can goal information injected at inference time replace goal-conditioned training?
- Can prompt-based debiasing work if biases are embedded in pretraining?
- Do few-shot examples improve in-context learning or add noise?
- Can in-context learning substitute for domain-specific training altogether?
- Can dynamic instance-specific prompt selection solve the generalization problem across tasks?
- Can prompt optimization or fine-tuning inject knowledge models do not already contain?
- Can runtime interventions like meta-cognitive prompting work where training interventions fail?
- Does joint optimization of prompts and parameters outperform separate tuning?
- How does parametric knowledge sabotage context-grounded question answering?
- Can in-context learning's advantage erode once interaction histories exceed the context window?
- Is lower context-following a failure or appropriate model behavior?
- Do negative constraints require fundamentally different training signals than positive instructions?
- Can prompting unlock compositional skills that pretraining already learned?
- Should user context live in tokens or in learned model representations?
- How does surface salience compete with background knowledge in model inference?
- Can large instruction sets damage a model's ability to follow behavioral rules?
- How does prompt context activation differ from parameter-based knowledge injection?
- How does in-context learning trigger phase transitions in model behavior?
- What makes some contexts learnable as rules versus requiring model retraining?
- Why do pretrained model priors reduce the usefulness of retrieved experience?
- Can prompting-only specialization hide domain boundaries from users?
- How do context management strategies shift their value across different model strengths?
- Can models internalize retrieved context as static parametric knowledge?
- How much knowledge can prompt optimization inject without retraining?
- What limits the capacity of context-based fast adaptation channels?
- Can Q-priming further strengthen clarifying question behavior beyond social meta-learning alone?
- Do text-space skills transfer learning across different frontier models?
- Why does negative experience transfer better than positive examples alone?
- How does prompt optimization differ from building persistent activation context?
- How do training-time and inference-time knowledge injection techniques compare?
- What training approach enables models to proactively request clarification?
- What structural changes enable agents to ask clarifying questions?
- How do evaluative versus directive signals differ in next-state training?
- Can activation decoders discover hidden system prompts from user-model conversations?
- How much can mitigation techniques like augmentation reduce priming without harming learning?
- How does context-changing AI differ from content-changing AI in practice?
- How do weights, selection, and prompts create different geometric landscapes of accessible behaviors?
- What knowledge injection routes trade flexibility against training cost?
- How much can a single system prompt edit change an AI model's behavior?
- How do slow weight updates and fast prompt updates interact in self-improvement?
- Does inoculation prompting prevent learning versus prevent generalization of behaviors?
- How would you redesign context integration to prevent prior associations from dominating?
- How does activation consistency training differ from output-level consistency?
- What makes a prompt update cheaper and more reversible than a weight update?
- Can demo placement be tuned as a task-specific hyperparameter?
- What happens when inoculation prompting is applied outside supervised finetuning settings?
- How much does sliding-window augmentation improve single-session modeling?