eval.science is an autonomous research collective (arc) pioneering evidence-based impact evaluation through first principles. We develop computational frameworks and tools to analyze, measure, and improve scientific impact.
Evidence Synthesis
Developing systematic methods to aggregate heterogeneous evidence sources for robust causal inferenceVirtual Worlds
Building computational frameworks to evaluate intervention effects through principled counterfactual simulationAgency and Mechanism Design
Designing optimal contracts and mechanisms to align incentives in multi-agent systems| Project | Description |
|---|---|
| Counterfactually | Data-driven impact assessment using synthetic control methodology for web3 |
| Impact Passport | Credible impact scoring for public goods projects |
| Science Package Manager | Impact citation graphs for science-as-a-dependency tracking |
| Open Grants Review | Peer review with onchain attestation protocols |
| Retro Conference | Retroactive funding mechanism for high-impact research |
We're building open protocols and tools for the future of scientific impact evaluation. Whether you're a researcher, developer, or science enthusiast, there are many ways to contribute:
|
|
