Research - Matthew Renze

Self-Reflection in LLM Agents: Effects on Problem-Solving Performance

Large-language model agents can reflect on their chain of thought to improve problem-solving performance.

Self-Reflection in LLM Agents: Effects on Problem-Solving Performance

The Effect of Sampling Temperature on Problem Solving in LLMs

Changing an LLM's sampling temperature from 0.0 to 1.0 does not affect problem-solving performance.

The Effect of Sampling Temperature on Problem Solving in LLMs

The Benefits of a Concise Chain-of-Thought on Problem-Solving in LLMs

A concise chain of thought in LLMs reduces verbosity and cost without impacting problem-solving.

The Benefits of a Concise Chain-of-Thought on Problem-Solving in LLMs

Curriculum Learning for RL Agents in a Pac-Man Environment

Curriculum learning improves the performance of some but not all RL agents in PacMan.

Curriculum Learning for RL Agents in a Pac-Man Environment

Natural Language Explanations for Explainable AI Systems using LLMs

LLMs can be used to create human-readable explanations for decisions made by AI systems.

Natural Language Explanations for Explainable AI Systems using LLMs

Automatically Generating Lecture Slides with Generative Pretrained Transformers

LLMs such as GPT-4 can be used to automate the creation of lecture slides from course material.

Automatically Generating Lecture Slides with Generative Pretrained Transformers

To see more of my research, please visit my Google Scholar and Semantic Scholar profiles.