Papers by Hadas Orgad
Diffusion Lens: Interpreting Text Encoders in Text-to-Image Pipelines (2024.acl-long)
Copied to clipboard
| Challenge: | Text-to-image diffusion models use a latent text prompt to guide image generation . however, the process by which the encoder produces the text representation is unknown . |
| Approach: | They propose a method for analyzing the text encoder of T2I models by generating images from its intermediate representations. |
| Outcome: | The proposed method provides valuable insights into the text encoder component in T2I pipelines. |
How Gender Debiasing Affects Internal Model Representations, and Why It Matters (2022.naacl-main)
Copied to clipboard
| Challenge: | Existing studies of gender bias in NLP focus on extrinsic or intrinsic bias, but the relationship between extrindic and intrinsic bias is relatively unknown. |
| Approach: | They propose a framework to measure extrinsic and intrinsic bias together and propose metric to measure debiasing and intrinsic debiases. |
| Outcome: | The proposed framework provides a comprehensive perspective on bias in NLP models, which can be applied to deploy NLP systems in a more informed manner. |
BLIND: Bias Removal With No Demographics (2023.acl-long)
Copied to clipboard
| Challenge: | Numerous methods to mitigate social biases require prior knowledge of the demographics in the dataset, such as gender or race. |
| Approach: | They propose a method for bias removal without prior knowledge of demographics in the dataset. |
| Outcome: | Experiments with racial and gender biases in sentiment classification and occupation classification tasks show that BLIND mitigates biase . BLINT is competitive with methods that require demographic information and sometimes surpasses them. |
Position-aware Automatic Circuit Discovery (2025.acl-long)
Copied to clipboard
| Challenge: | Existing circuit analysis methods treat component positions as equally relevant across input positions, limiting their ability to capture cross-positional interactions or mechanisms that vary across positions. |
| Approach: | They propose to extend edge attribution patching to differentiate between token positions and introduce a dataset schema to enable position-aware circuit discovery in datasets with variable length examples. |
| Outcome: | The proposed approach enables position-aware circuit discovery even on tasks with variable length examples. |
ReFACT: Updating Text-to-Image Models by Editing the Text Encoder (2024.naacl-long)
Copied to clipboard
| Challenge: | Text-to-image generative models encode factual associations that can quickly become outdated, diminishing their utility for end-users. |
| Approach: | They propose a method for editing factual associations in text-to-image models without retraining or explicit input from end-users. |
| Outcome: | The proposed method improves generalization and preservation of unrelated concepts on an existing dataset and compares with other methods. |
Padding Tone: A Mechanistic Analysis of Padding Tokens in T2I Models (2025.naacl-long)
Copied to clipboard
| Challenge: | Text-to-image (T2I) diffusion models rely on encoded prompts to guide the image generation process. |
| Approach: | They conduct the first in-depth analysis of the role padding tokens play in T2I diffusion models by using two causal techniques to analyze how information is encoded in the representation of tokens across different components of the pipeline. |
| Outcome: | The proposed techniques reveal that padding tokens may affect the model’s output during text encoding, during the diffusion process, or be effectively ignored. |