How does this skill help identify 'good' research?

It uses a structured analysis of landmark papers to find commonalities such as introducing new primitives, creating surprising connections, or providing rigorous foundations to existing folklore.

What is research taste?

Research taste is the pattern-recognition ability to distinguish work that will fundamentally change a field from work that is merely incremental, even before the community reaches a consensus.

Can this skill help me avoid low-impact work?

Yes, it specifically identifies 'Anti-Patterns' like the Incremental Treadmill, Method Mashing, and the Complexity Monster that typically result in work with low long-term utility.

Is this skill only for machine learning researchers?

While the examples provided are often from the AI/ML community, the underlying principles of problem selection, execution taste, and communication apply to almost any scientific or technical research domain.

What are the exercises included in this skill?

The skill includes the Prediction Game (forecasting citations), Explain the Gap (comparing highly-cited vs. low-cited work), and The Time Machine (historical context analysis).

Research Taste Developer

Name: Research Taste Developer
Author: GhostScientist

byGhostScientist

•

Learning & Documentation

Develops acute intuition for identifying and creating high-impact research by analyzing patterns in landmark papers and avoiding incremental pitfalls.

About

Research Taste Developer provides a structured framework to help researchers distinguish significant work from incremental progress. By analyzing highly-cited landmark papers, identifying impact patterns like new primitives or surprising connections, and flagging common anti-patterns like method mashing or benchmark overfitting, this skill guides users through developing the pattern recognition necessary to pick important problems. It offers actionable exercises to calibrate intuition, helping you predict which ideas will shape the future of a field versus those that will be forgotten.

Key Features

Generates structured Taste Development Reports for specific academic subfields
Provides deliberate practice exercises to calibrate research prediction and intuition
Analyzes landmark papers to extract core insights and high-impact patterns
Identifies common research anti-patterns like the incremental treadmill and complexity monsters
1 GitHub stars
Categorizes successful work into archetypes such as New Primitives and Scaling Insights

Use Cases

Onboarding into a new academic subfield by identifying truly foundational literature
Evaluating a research project's potential impact and longevity before committing resources
Improving paper communication by focusing on the core insight rather than just the method

About

Key Features

Generates structured Taste Development Reports for specific academic subfields
Provides deliberate practice exercises to calibrate research prediction and intuition
Analyzes landmark papers to extract core insights and high-impact patterns
Identifies common research anti-patterns like the incremental treadmill and complexity monsters
1 GitHub stars
Categorizes successful work into archetypes such as New Primitives and Scaling Insights

Use Cases

Onboarding into a new academic subfield by identifying truly foundational literature
Evaluating a research project's potential impact and longevity before committing resources
Improving paper communication by focusing on the core insight rather than just the method