Posts – Latent Space Notes

Evaluating o1-preview for prompt injection

AI In Action

LLMs

Evals

Baruch iteratively tests different prompt versions using OpenAI’s new o1-preview model, analyzing results and refining the prompts to identify potential security risks.

Validating The LLM Validators with Shreya

Paper Club

LLMs

Evals

We discuss the challenges of LLM validators and emphasizing the iterative nature of defining good evaluation criteria and aligning LLMs to those criteria.

AI Powered IDE Alternatives (Part 2)

AI In Action

Dev Tools

IDE

Exploring Melty, Void and Aider as alternatives to Cursor, the AI-powered code editor. The session covers comparative analysis for open-source alternatives in the rapidly…

LLM Reasoning: Q-STaR and Friends

Paper Club

Reasoning

LLMs

This week’s Paper Club is focused on LLM Reasoning from the arXiv papers STaR, Quiet-STaR, and V-STaR.

AI Powered IDE Alternatives (Part 1)

AI In Action

Dev Tools

IDE

An exploration of IDEs, featuring Cursor, PearAI and NeoVim - AI-powered coding tools. The session covers comparative analysis, practical demonstrations, and open-source…

Long Context Retrieval by Writer

Paper Club

Retrieval

LLMs

We explore Writing in the Margins, a technique for improving long context retrieval in LLMs. Learn how this method enhances performance without fine-tuning, making it easier…

Umar Jamil, Sam Julien

Langflow: A Visual LLM Tool

AI In Action

Dev Tools

LLMs

A presentation about Langflow, an open-source tool for building and managing LLM-powered applications using a visual node-based interface.

AI Research Agents: Storm, Scientist, and GPTR

AI In Action

Agents

This week’s AI In Action club is focused on AI research agents, specifically three tools.

Yikes , Frikster