Turn a research anchor into auditable candidate evidence.
deep learning research
Explores bounded deep-learning research candidates against a durable research anchor while preserving auditable, reproducible comparisons.
When to use it
Use when the researcher explicitly authorizes exploration and supplies a durable current-research context.
Give it an authorized research campaign in a project; it evaluates candidates and writes candidate-only evidence into the project.
What you provide
This skill
Zotero
Reads your Zotero literature
arXiv API
Fetches metadata from arXiv
~/.rigorpilot/PERSONAL_RIGOR.md
Reads your personal rigor settings
Python is needed to use the deterministic orchestration and output scripts.
Requires shared operating, research-rigor, experiment, and thinking-loop references from ai-research-reproduction.
An available Zotero library enables preferred local curated-literature lookup.
Usually delegates construction of campaign-specific repository-understanding artifacts to analyze-project.