Papers by Kira Radinsky
Leveraging Prototypical Representations for Mitigating Social Bias without Demographic Information (2024.naacl-short)
Copied to clipboard
| Challenge: | Existing approaches to mitigate social biases require explicit annotation of demographic information for each sample. |
| Approach: | They propose a method that leverages predefined demographic texts and incorporates a regularization term during the fine-tuning process to mitigate bias in language models. |
| Outcome: | The proposed method outperforms debiasing methods with limited demographic-annotated data. |
Cross-Cultural Transfer Learning for Text Classification (D19-1)
Copied to clipboard
| Challenge: | a large dataset is required to achieve competitive performance in most natural language tasks. large datasets are expensive, time consuming, and error-prone. |
| Approach: | They propose a transfer-learning framework that leverages bilingual corpora for natural language text classification using no task-specific data. |
| Outcome: | The proposed framework can achieve good performance on formality classification and sarcasm detection tasks without any task-specific labeled data. |
Propaganda Signals in LLMs: Perspectival Divergence and Narrative Framing in the Russia-Ukraine War (2026.findings-acl)
Copied to clipboard
| Challenge: | Large Language Models are increasingly used to explain, summarize, and translate real-world events . a recent study examined whether LLMs reproduce conflict-specific propaganda . |
| Approach: | They evaluate LLMs under several prompting contexts to determine which side they are closer to . they find model-specific leanings and technique profiles that persist across prompts . |
| Outcome: | The proposed model outputs align with competing narratives from different information ecosystems. |
Shielded Representations: Protecting Sensitive Attributes Through Iterative Gradient-Based Projection (2023.findings-acl)
Copied to clipboard
| Challenge: | Natural language processing models tend to learn and encode social biases present in the data. |
| Approach: | They propose a method for removing non-linear encoded concepts from neural representations by iteratively training neural classifiers to predict a particular attribute, followed by a projection of the representation on a hypersurface. |
| Outcome: | The proposed method removes non-linear encoded concepts from neural representations. |
Temporal Attention for Language Models (2022.findings-naacl)
Copied to clipboard
| Challenge: | Pretrained language models are trained on corpora derived from the web, but ignore this information. |
| Approach: | They propose a time-aware self-attention mechanism that captures time-specific contextualized word representations and allows the transformer to capture this information. |
| Outcome: | The proposed model achieves state-of-the-art on three datasets in different languages (English, German, and Latin) that vary in time, size, and genre. |
Clinical Contradiction Detection (2023.emnlp-main)
Copied to clipboard
| Challenge: | Detecting contradictions in text is difficult since it requires clinical expertise. |
| Approach: | They propose to use a medical ontology to build a seed of potential medical contradictions in medical abstracts by distant supervision. |
| Outcome: | The proposed method weakly supervises state-of-the-art deep learning models and shows significant improvements across multiple medical contradiction datasets. |