Sundial Org

This skill evaluates ML research papers following ICML guidelines, providing comprehensive, literature-grounded feedback across claims, methods, experiments,

This skill guides you through tree-based research cycles as an AI co-scientist, ensuring hypothesis-driven, reproducible experiments across domains.

This skill helps you estimate Tinker fine-tuning costs by tokenizing datasets with the correct tokenizer and applying model-specific pricing.

This skill critiques ML conference papers in reviewer style, helping you anticipate concerns, identify weaknesses, and fill claim-evidence gaps.

This skill guides ideation and implementation of neuro-symbolic reasoning, combining LLMs with symbolic solvers to design experiments or integrate solvers.

This skill conducts a literature review and crafts a CS research proposal by identifying assumptions, proposing alternatives, and outlining testable validation.

This skill helps you split large uncommitted changes into logical commits by proposing a safe, numbered plan and applying patches precisely.