Human-Centered Disability Bias Detection in Large Language Models
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
To promote a more just and inclusive society, developers and researchers are strongly encouraged to design Language Models (LM) with ethical considerations at the forefront, ensuring that the benefits and opportunities of AI are accessible to all users and communities.Incorporating humans in the loop is one approach recognized for mitigating general AI biases.Consequently, the development of new design guidelines and datasets is essential to help AI systems realize their full potential for the benefit of people with disabilities.This study aims to identify disability-related bias in Large Masked Language Models (MLMs), the Electra.A participatory and collaborative research approach was employed, involving three disability organizations to collect information on deaf and hard-of-hearing individuals.Our initial analysis reveals that the studied MLM is highly sensitive to the various identity references used to describe deaf and hard-of-hearing people.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it