6 papers
Geometry of Divergence: Tracking Hidden-State Trajectories for Adaptive Multi-Turn Reasoning
Jie Liang, Zhengxin Yu, Hamid Nasiri +1
LLM agents need to sustain goal-consistent reasoning across long multi-turn interactions under strict resource constraints. However, as the multi-turn context accumulates, it can d…
kNNGuard: Turning LLM Hidden Activations into a Training-Free Configurable Guardrail
Mahmoud Abdelfattah, Hamid Nasiri, Peter Garraghan
Large language models (LLMs) are increasingly deployed in domains requiring guardrails to detect unsafe, off-topic, or adversarial prompts. Existing guardrails predominantly rely o…
Cost-Aware Model Orchestration for LLM-based Systems
Daria Smirnova, Hamid Nasiri, Marta Adamska +2
As modern artificial intelligence (AI) systems become more advanced and capable, they can leverage a wide range of tools and models to perform complex tasks. The task of orchestrat…
Green Prompting: Characterizing Prompt-driven Energy Costs of LLM Inference
Marta Adamska, Daria Smirnova, Hamid Nasiri +2
Large Language Models (LLMs) have become widely used across various domains spanning search engines, code generation, and text creation. However, a major concern associated with th…
EDoRA: Efficient Weight-Decomposed Low-Rank Adaptation via Singular Value Decomposition
Hamid Nasiri, Peter Garraghan
Parameter-efficient fine-tuning methods, such as LoRA, reduces the number of trainable parameters. However, they often suffer from scalability issues and differences between their…
ICT Sector Greenhouse Gas Emissions -- Issues and Trends
Peter Garraghan, John Hutchinson, Adrian Friday
As Information and Communication Technology (ICT) use has become more prevalent, there has been a growing concern in how its associated greenhouse gas emissions will impact the cli…