Summary of Large Language Model Confidence Estimation Via Black-box Access, by Tejaswini Pedapati et al.
Large Language Model Confidence Estimation via Black-Box Access
by Tejaswini Pedapati, Amit Dhurandhar, Soumya Ghosh, Soham Dan, Prasanna Sattigeri
First submitted to arxiv on: 1 Jun 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
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 framework estimates confidence in responses from large language models (LLMs) using simple and extensible methods. The approach engineers novel features and trains an interpretable model to predict confidence, demonstrating effectiveness on multiple benchmark tasks. For instance, the confidence models built for Flan-ul2 and Llama-13b generalize well across different datasets, providing insights into predictive features. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Large language models (LLMs) are great tools, but it’s hard to know how much you can trust their answers. This paper helps with that by showing a simple way to figure out how confident an LLM is in its responses. They do this by creating new “features” that help train a model to predict confidence. This works really well on lots of different tasks, and it’s even better when they use the same approach for multiple models. |
Keywords
» Artificial intelligence » Llama
