Call for Papers

What we are looking for

We welcome empirical negative results, unexpected model behavior, and careful analyses of distribution shift, confounding, weak supervision, resource constraints, benchmark misalignment, uncertainty, causal learning, and adaptive strategies in biological applications.

Full papers

Full submissions may use up to eight pages, excluding references and appendices, and should include:

• A clear biological task, data modality, and modeling approach.
• Empirical or theoretical evidence of a negative or unexpected outcome.
• Analysis of the model, data, experimental, or deployment causes.
• Reproducible and well-documented results.

Tiny papers

Tiny papers may use up to four pages. They must state the problem and provide empirical evidence of at least one negative outcome, even when the full causal analysis is still developing.

Formatting and review

• Use the NeurIPS 2026 LaTeX style.
• Submissions are double-blind; linked material must also preserve anonymity.
• Accepted papers will appear on OpenReview, and the workshop remains non-archival.
• Concurrent submissions are welcome when they comply with the target venue’s policy.

Eligibility

Submissions are considered only if the work has not been accepted for publication in previous conference proceedings.

Evaluation criteria

• Clarity of the problem and claims.
• Technical rigour and reproducibility.
• Faithfulness to the biological setting.
• Depth of failure analysis.
• Quality of empirical documentation.

Tentative dates

Submission

August 29, 2026

Review

August 29–September 21, 2026

Notification

September 29, 2026

Camera-ready & poster

October 20, 2026

All deadlines are 11:59 p.m. Anywhere on Earth.