Summary of From Words to Numbers: Your Large Language Model Is Secretly a Capable Regressor When Given In-context Examples, by Robert Vacareanu et al.
From Words to Numbers: Your Large Language Model Is Secretly A Capable Regressor When Given In-Context Examples
by Robert Vacareanu, Vlad-Andrei Negru, Vasile Suciu, Mihai Surdeanu
First submitted to arxiv on: 11 Apr 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
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 Large pre-trained language models, such as GPT-4, Claude 3, and Llama2, can perform linear and non-linear regression tasks using in-context examples without additional training. Our research shows that these models rival or outperform traditional supervised methods like Random Forest, Bagging, and Gradient Boosting on certain datasets. For example, Claude 3 outperformed many supervised methods on the Friedman #2 regression dataset. We also investigated how well large language model performance scales with the number of in-context exemplars, borrowing from online learning concepts to demonstrate that LLMs can achieve sub-linear regret. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Large pre-trained language models are super smart! They can do a job called “regression” which is like trying to predict a value based on some information. The researchers tested these models and found out they’re really good at it! Some of them even did better than other traditional methods that experts use. One model, Claude 3, was especially good at a tricky task called Friedman #2 regression dataset. The scientists also looked into how well these models do as you give them more information to work with, and they found out that the models can get even better! |
Keywords
» Artificial intelligence » Bagging » Boosting » Claude » Gpt » Large language model » Linear regression » Online learning » Random forest » Regression » Supervised
