Rachel M. Harrison

Rachel M. Harrison

Random Number Generation Tasks (RNGTs) are used in psychology for examining how humans generate sequences devoid of predictable patterns. By adapting an existing human RNGT for an LLM-compatible environment, this preliminary study tests whether ChatGPT-3.5, a large language model (LLM) trained on human-generated text, exhibits human-like cognitive biases when generating random number sequences. Initial findings indicate that ChatGPT-3.5 more effectively avoids repetitive and sequential patterns compared to humans, with notably lower repeat frequencies and adjacent number frequencies. Continued research into different models, parameters, and prompting methodologies will deepen our understanding of how LLMs can more closely mimic human random generation behaviors, while also broadening their applications in cognitive and behavioral science research.

Daniel Grimes, Rachel M. Harrison

This paper presents BLK-Assist, a modular framework for artist-specific fine-tuning of diffusion models using parameter-efficient methods. The system is implemented as a case study with a single professional artist's proprietary corpus and consists of three components: BLK-Conceptor (LoRA-adapted conceptual sketch generation), BLK-Stencil (LayerDiffuse-based transparency-preserving asset generation), and BLK-Upscale (hybrid Real-ESRGAN and texture-conditioned diffusion for high-resolution outputs). We document dataset composition, preprocessing, training configurations, and inference workflows to enable reproducibility with publicly available models to illustrate a privacy-preserving, consent-based approach to human-AI co-creation that maintains stylistic fidelity to the source corpus and can be adapted for other artists under similar constraints.

Oluwole Fagbohun, Rachel M. Harrison, Anton Dereventsov

Due to rapid advancements in the development of Large Language Models (LLMs), programming these models with prompts has recently gained significant attention. However, the sheer number of available prompt engineering techniques creates an overwhelming landscape for practitioners looking to utilize these tools. For the most efficient and effective use of LLMs, it is important to compile a comprehensive list of prompting techniques and establish a standardized, interdisciplinary categorization framework. In this survey, we examine some of the most well-known prompting techniques from both academic and practical viewpoints and classify them into seven distinct categories. We present an overview of each category, aiming to clarify their unique contributions and showcase their practical applications in real-world examples in order to equip fellow practitioners with a structured framework for understanding and categorizing prompting techniques tailored to their specific domains. We believe that this approach will help simplify the complex landscape of prompt engineering and enable more effective utilization of LLMs in various applications. By providing practitioners with a systematic approach to prompt categorization, we aim to assist in navigating the intricacies of effective prompt design for conversational pre-trained LLMs and inspire new possibilities in their respective fields.

Rachel M. Harrison, Anton Dereventsov, Anton Bibin

We present a method for zero-shot recommendation of multimodal non-stationary content that leverages recent advancements in the field of generative AI. We propose rendering inputs of different modalities as textual descriptions and to utilize pre-trained LLMs to obtain their numerical representations by computing semantic embeddings. Once unified representations of all content items are obtained, the recommendation can be performed by computing an appropriate similarity metric between them without any additional learning. We demonstrate our approach on a synthetic multimodal nudging environment, where the inputs consist of tabular, textual, and visual data.