R0049/2026-03-31/Q003-SRC01 — Scorecard¶
Source¶
| Title | PaperQA2 — High accuracy RAG for answering questions from scientific documents |
| Publisher | FutureHouse / GitHub |
| Authors | FutureHouse team |
| Date | 2024-2025 |
| URL | https://github.com/Future-House/paper-qa |
| Type | Open-source software repository |
Ratings¶
| Dimension | Rating |
|---|---|
| Reliability | High |
| Relevance | High |
| Missing data | Low risk |
| Measurement | Low risk |
| Selective reporting | Low risk |
| Randomization | N/A |
| Protocol deviation | N/A |
| COI/funding | Low risk |
Rationale¶
| Dimension | Rationale |
|---|---|
| Reliability | Open-source, peer-reviewed academic paper, benchmarked against humans |
| Relevance | Most advanced open-source academic research agent — a direct comparator for structured analytical features |
Evidence Extracts¶
| Evidence | Summary |
|---|---|
| SRC01-E01 | PaperQA2 achieves superhuman scientific Q&A but implements none of the five target analytical features |