prompt engineering
This process involves designing effective input prompts for generative models, particularly in natural language processing, to elicit desired responses or outputs. It requires understanding the behavior of models so as to maximize the relevance and coherence of the generated content.
- ConfTuner: Training Large Language Models to Express Their Confidence Verbally
- Do LLMs Really Forget? Evaluating Unlearning with Knowledge Correlation and Confidence Awareness
- Don’t Forget the Enjoin: FocalLoRA for Instruction Hierarchical Alignment in Large Language Models
- From Indicators to Insights: Diversity-Optimized for Medical Series-Text Decoding via LLMs
- MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models
- No Loss, No Gain: Gated Refinement and Adaptive Compression for Prompt Optimization
- Quantifying Elicitation of Latent Capabilities in Language Models
- Short-length Adversarial Training Helps LLMs Defend Long-length Jailbreak Attacks: Theoretical and Empirical Evidence
- The Fragile Truth of Saliency: Improving LLM Input Attribution via Attention Bias Optimization
- Treasure Hunt: Real-time Targeting of the Long Tail using Training-Time Markers