🐙 Lunch & Learn: LLM & Prompting 101

TL;DR

Interactive session on LLMs, prompting, and a scavenger hunt.

Transcript

hey friends sorry I am late by five minutes yes for five minutes I apologize for that how everybody doing hey Victor hey adash me actually refresh this I just want to see is it working I don't think it's actually working wait oh no it is never mind it is working um it's not starting it will start soon yes it has started now so I apologize for this ... Read More

Key Insights

• The session involves learning about large language models (LLMs) through an interactive scavenger hunt, enhancing understanding of how LLMs function.
• Participants are encouraged to engage in prompt engineering exercises, which involve generating specific outputs from LLMs using carefully crafted prompts.
• The scavenger hunt includes solving challenges related to generating JSON data, understanding context windows, and controlling creativity in LLMs.
• A key part of the session is understanding the importance of context in LLMs, demonstrated through exercises with unigram, bigram, and trigram models.
• The session highlights the use of Brilliant.org for interactive learning, with modules focused on LLMs and STEM subjects.
• Participants are encouraged to complete additional modules to reveal a five-letter word, related to a dragon-type Pokémon with aspirations to fly.
• The session is sponsored by Brilliant.org, offering interactive courses and a discount for participants who sign up through the provided link.
• The session concludes with a light-hearted challenge where participants can request a song from the host if they solve the scavenger hunt.

Questions & Answers

Q: What is the main focus of the session?

The main focus of the session is to provide an interactive learning experience about large language models (LLMs) and prompt engineering. Participants engage in a scavenger hunt challenge and exercises to understand context windows, JSON data generation, and controlling creativity in LLMs.

Q: How is the scavenger hunt structured?

The scavenger hunt is structured around solving challenges related to LLMs, such as generating JSON data and understanding context windows. Participants complete modules on Brilliant.org and LeetLLM to reveal a five-letter word related to a dragon-type Pokémon with aspirations to fly.

Q: What platforms are used during the session?

The session utilizes platforms like Brilliant.org and LeetLLM. Brilliant.org offers interactive STEM courses, while LeetLLM provides a platform for learning prompt engineering through Q&A challenges. Participants use these platforms to complete exercises and challenges related to LLMs.

Q: What is the role of prompt engineering in the session?

Prompt engineering plays a crucial role in the session as participants learn to generate specific outputs from LLMs using carefully crafted prompts. Exercises include generating JSON data and understanding context windows, demonstrating the importance of precise prompts in interacting with LLMs.

Q: What educational resources are recommended in the session?

The session recommends completing Brilliant.org's modules on LLMs and STEM subjects. Participants are encouraged to finish modules related to context windows, controlling creativity, and improving models to enhance their understanding of LLMs and complete the scavenger hunt challenge.

Q: What is the significance of the scavenger hunt's five-letter word?

The five-letter word is part of the scavenger hunt challenge and is related to a dragon-type Pokémon known for its aspirations to fly. Solving the challenge involves completing modules and exercises related to LLMs, reinforcing the learning objectives of the session.

Q: What incentives are offered for participating in the session?

Participants who complete the session's challenges and exercises can receive a 20% discount on Brilliant.org's annual membership. Additionally, if participants solve the scavenger hunt, they can request a song from the host, adding a fun and engaging element to the session.

Q: How does the session conclude?

The session concludes with a light-hearted challenge where participants who solve the scavenger hunt can request a song from the host. This adds an entertaining and interactive element to the learning experience, encouraging participants to engage fully with the exercises and challenges.

Summary & Key Takeaways

• The session is an interactive learning experience focused on large language models (LLMs) and prompt engineering, featuring a scavenger hunt challenge.

• Participants engage in exercises to generate JSON data and learn about context windows in LLMs, using platforms like Brilliant.org and LeetLLM.

• The session is sponsored by Brilliant.org, offering participants a chance to explore interactive STEM courses and receive a discount on membership.