4 results for "prompt optimization"
CAP-CoT: Cycle Adversarial Prompt for Improving Chain of Thoughts in LLM Reasoning
Chain-of-Thought (CoT) prompting has emerged as a simple and effective way to elicit step-by-step solutions from large language models (LLMs). However, CoT reasoning can be unstable across runs on lon…
Discovering Agentic Safety Specifications from 1-Bit Danger Signals
Can large language model agents discover hidden safety objectives through experience alone? We introduce EPO-Safe (Experiential Prompt Optimization for Safe Agents), a framework where an LLM iterative…
Agentic Adversarial Rewriting Exposes Architectural Vulnerabilities in Black-Box NLP Pipelines
Multi-component natural language processing (NLP) pipelines are increasingly deployed for high-stakes decisions, yet no existing adversarial method can test their robustness under realistic conditions…
Intel B70: LLama.ccp SYCL vs LLama.cpp OpenVino vs LLM-Scaler
In case anyone is interested, I decided to test out LLama.cpp's new OpenVino backend to see how it compares on Intel GPUs. At first glance, it stomps all over the previous best-case, SYCL, but lags be…