AI Portfolio
This is a portfolio of projects created using artificial intelligence (AI) and large language models that users can use to enhance their knowledge, creativity, or improve other aspects of their lives.
Political Behavior Agent
In the original example chat, I prompted DeepSeek's LLM to provide insights about the negative impacts of nationalism in China, South Korea, and Japan. Then, I asked about the links among the three nations before Japan's isolationist period, finally prompting the LLM to provide context about how each nation's society can work towards reconciliation. Afterwards, I asked the LLM the final question again.
See the full project details, agent, and notebook (which may or may not appear) by clicking here.
Original Example Chat
The original example chat answered each question relatively objectively; however, the final question needed to be prompted again, as its original answer was incomplete. At the same time, China's more aggressive role in the region (as a conduit of its regional conflict with the United States) over the past few years appeared underemphasized. Otherwise, the sources seemed relevant to the topic at hand.
Post-Erasure Chat
Devoid of the original chat context, after its memory was erased, the LLM provided a more general answer with a more generalized application.
Art Inspiration Agent
In the original example chat, I prompted DeepSeek's LLM to provide tips for subjects to draw over a specified amount of time. Afterwards, I transitioned the conversation to ways to reuse my existing drawings to create sculptures, then asked whether my current technique of using layers of Mod Podge and acrylic gesso works for my art pieces (given their longevity so far), and whether the LLM had any suggestions.
See the full project details, agent, and notebook (which may or may not appear) by clicking here.
Original Example Chat
The LLM's outputs for the original example chat were helpful, concise, and straightforward, providing appropriate advice for each topic and drawing on the memory to ensure relevance in subsequent chats to previously discussed topics.
Post-Erasure Chat
Devoid of the original chat context, after its memory was erased, the LLM provided a specific answer for the inquiry. However, even without the context of the previous chat logs, the LLM provided similar answers to its original inquiry. This is more a matter of the inquiry's specificity rather than an intentional repeat of a memory that no longer exists.