Papers by Inderjit Dhillon
Extreme Zero-Shot Learning for Extreme Text Classification (2022.naacl-main)
Copied to clipboard
| Challenge: | Experimental results show that MACLR achieves superior performance compared to other baseline methods. |
| Approach: | They propose to pre-train Transformer-based encoders with self-supervised contrastive losses to learn the semantic embeddings of instances and labels with raw text. |
| Outcome: | The proposed method improves on the EZ-XMC model with a limited number of ground-truth positive pairs. |
Automatic Engineering of Long Prompts (2024.findings-acl)
Copied to clipboard
| Challenge: | Recent research has explored automatic prompt engineering for short prompts, typically consisting of one or a few sentences. |
| Approach: | They propose an algorithm that automatically improves long prompts by combining a greedy algorithm with beam-search to enhance the effectiveness of LLM-based mutation. |
| Outcome: | The proposed algorithm achieves 9.2% accuracy gain on eight tasks in Big Bench Hard and consistent improvements on GSM8K with various models. |
PRISM: A New Lens for Improved Color Understanding (2024.emnlp-industry)
Copied to clipboard
| Challenge: | PRISM is a visual representation learner that can grasp the nuances of precise colors without compromising CLIP’s performance on established benchmarks. |
| Approach: | They propose a method that extends CLIP's ability to grasp the nuances of precise colors by utilizing a curated dataset of 100 image-text pairs that can be effortlessly repurposed for fine-tuning. |
| Outcome: | The proposed method improves CLIP's ability to grasp the nuances of precise colors without compromising CLIP’s performance on established benchmarks. |