Summary of Procedural Adherence and Interpretability Through Neuro-symbolic Generative Agents, by Raven Rothkopf et al.
Procedural Adherence and Interpretability Through Neuro-Symbolic Generative Agents
by Raven Rothkopf, Hannah Tongxin Zeng, Mark Santolucito
First submitted to arxiv on: 24 Feb 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Machine Learning (cs.LG); Logic in Computer Science (cs.LO)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed approach combines formal logic-based program synthesis and large language model (LLM) content generation to create interactive agents that guarantee procedural adherence and interpretability. This is achieved by using Temporal Stream Logic (TSL) to generate an automaton that enforces a high-level temporal structure on the agent, allowing it to focus on a shorter context window. The approach was evaluated on tasks for creating choose-your-own-adventure games, resulting in at least 96% adherence to temporal constraints with the automaton-enhanced agent compared to as low as 14.67% with the purely LLM-based agent. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine having an AI that can help create interactive stories or games. Right now, these systems are not very good at following rules and making sense of what’s happening over time. The researchers came up with a new way to make AI agents work better by combining two different approaches: formal logic and language models. This helps the AI agent make decisions and follow rules while still being able to generate creative content. They tested this idea on creating choose-your-own-adventure games and found that their approach worked much better than just using language models alone. |
Keywords
* Artificial intelligence * Context window * Large language model
