Summary of Explorer: Exploration-guided Reasoning For Textual Reinforcement Learning, by Kinjal Basu et al.
EXPLORER: Exploration-guided Reasoning for Textual Reinforcement Learning
by Kinjal Basu, Keerthiram Murugesan, Subhajit Chaudhury, Murray Campbell, Kartik Talamadupula, Tim Klinger
First submitted to arxiv on: 15 Mar 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); 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 A neurosymbolic reinforcement learning agent, called EXPLORER, is introduced for textual reinforcement learning. The agent combines a neural module for exploration with a symbolic module for exploitation, allowing it to learn generalized symbolic policies that perform well on unseen data. This approach outperforms baseline agents in Text-World cooking (TW-Cooking) and Text-World Commonsense (TWC) games. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Text-based games are an important part of NLP tasks, requiring reinforcement learning agents to understand language and reason. The challenge is generalizing across multiple games and performing well on both seen and unseen objects. Some approaches work well on seen objects but fail on unseen ones. Others may work better on unseen data but their policies are not interpretable or easily transferable. To solve this problem, a new agent called EXPLORER was created. It’s like a special computer program that uses words to explore and make decisions. This program can learn new things and do well even when it doesn’t know something before. |
Keywords
» Artificial intelligence » Nlp » Reinforcement learning
