Summary of Cicle: Conformal In-context Learning For Largescale Multi-class Food Risk Classification, by Korbinian Randl et al.
CICLe: Conformal In-Context Learning for Largescale Multi-Class Food Risk Classification
by Korbinian Randl, John Pavlopoulos, Aron Henriksson, Tony Lindgren
First submitted to arxiv on: 18 Mar 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: 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 This paper presents a dataset of 7,546 short texts describing public food recall announcements, with manual labels for food products and hazards. The authors benchmark traditional and transformer-based models on this dataset, finding that logistic regression outperforms RoBERTa and XLM-R on classes with low support. Additionally, the paper discusses prompting strategies and introduces an LLM-in-the-loop framework based on conformal prediction, which improves performance while reducing energy consumption. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This research helps keep food safe by teaching computers to automatically detect when food is contaminated or spoiled. The scientists created a big database of short texts about food recalls and labeled them for specific types of products and hazards. They tested different computer models to see how well they could recognize these labels, and found that a simple approach worked better than more advanced ones on certain tasks. The researchers also showed how to make this process more efficient by using special prompts and techniques. |
Keywords
* Artificial intelligence * Logistic regression * Prompting * Recall * Transformer
