Machine learning models explanations as interpretations of evidence: a theoretical framework of explainability and its implications on high-stakes biomedical decision-making
Bibliographic record
Abstract
Explainable Artificial Intelligence, or XAI, is a vibrant research topic in the artificial intelligence community. It is raising growing interest across methods and domains, especially those involving high-stakes decision-making, such as the biomedical sector. Much has been written about the subject, yet XAI still lacks shared terminology and a framework capable of providing structural soundness to explanations, a crucial need for decisions that impact healthcare. In our work, we address these issues by proposing a novel definition of explanation that synthesizes insights from the existing literature. We recognize that explanations are not atomic, but rather the combination of evidence stemming from the model and its input-output mapping, along with the human interpretation of this evidence. Furthermore, we fit explanations into the properties of faithfulness (i.e., the explanation is an accurate description of the model’s inner workings and decision-making process) and plausibility (i.e., how much the explanation seems convincing to the user). Our theoretical framework simplifies the operationalization of these properties and provides new insights into common explanation methods that we analyze through case studies. We explore the impact of our framework in the sensitive domain of biomedicine, where XAI can play a central role in generating trust by balancing faithfulness and plausibility.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.025 | 0.082 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.006 | 0.003 |
| Science and technology studies | 0.002 | 0.019 |
| Scholarly communication | 0.006 | 0.013 |
| Open science | 0.004 | 0.006 |
| Research integrity | 0.004 | 0.004 |
| Insufficient payload (model declined to judge) | 0.006 | 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".