Summary of Insaaf: Incorporating Safety Through Accuracy and Fairness | Are Llms Ready For the Indian Legal Domain?, by Yogesh Tripathi et al.
InSaAF: Incorporating Safety through Accuracy and Fairness | Are LLMs ready for the Indian Legal Domain?
by Yogesh Tripathi, Raghav Donakanti, Sahil Girhepuje, Ishan Kavathekar, Bhaskara Hanuma Vedula, Gokul S Krishnan, Shreya Goyal, Anmol Goel, Balaraman Ravindran, Ponnurangam Kumaraguru
First submitted to arxiv on: 16 Feb 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 Medium Difficulty summary: This research paper investigates the ability of Large Language Models (LLMs) to perform legal tasks in the Indian landscape when social factors are involved. The study proposes a novel metric, -weighted Legal Safety Score (LSS_{}), which evaluates both the fairness and accuracy aspects of LLMs. The authors assess LLMs’ safety by considering their performance in the Binary Statutory Reasoning task and their fairness exhibition with respect to various axes of disparities in Indian society. They find that the proposed LSS_{} metric can effectively determine the readiness of a model for safe usage in the legal sector. Additionally, the authors propose finetuning pipelines using specialized legal datasets as a potential method to mitigate bias and improve model safety. The study’s results demonstrate that LLMs can be trained to perform well on legal tasks while reducing bias and improving their overall safety. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Low Difficulty summary: This paper explores how artificial intelligence language models can help with legal tasks in India. It shows that these models can learn biases and make unfair predictions, which is a problem. The researchers created a new way to measure the fairness and accuracy of these models, called the Legal Safety Score. They tested this score on two types of language models and found that it works well. The study also suggests ways to improve the fairness and safety of these models by fine-tuning them with specialized legal data. |
Keywords
» Artificial intelligence » Fine tuning
