BLADE Decision-Discrimination MCQ
The BLADE track for selecting the most or least justifiable conceptual-variable and data-transformation decisions for a research question and dataset.
Scientific task · Scientific workflow
Scientific evidence interpretation
Interpret or reconcile scientific text
科学证据解读
Definition and search aliases
- Permanent ID
scientific-evidence-interpretation- Aliases
- figure interpretation, table interpretation, literature synthesis
- Deprecated aliases
- None
- Hierarchy
- Leaf task under Scientific workflow and systems analysis
Coverage
2 benchmark families cover this task
LAB-Bench
A practical biology-research suite of 2,457 multiple-choice questions across eight broad categories and 31 versioned task files, with public and private contamination-monitoring splits.
LAB-Bench FigQA
Multiple-choice interpretation and multi-element reasoning over scientific figures shown without captions or paper context.
LAB-Bench LitQA2
Literature-retrieval questions whose answers require findings in full research papers rather than titles or abstracts.
LAB-Bench SuppQA
Retrieval and interpretation questions answerable from paper supplementary text or PDF tables.
LAB-Bench TableQA
Lookup, calculation, and reasoning questions over table images extracted from scientific papers.
Evidence-backed count claims
Each row keeps its original unit and basis. Rows with different units or overlapping mappings are never added.
| Benchmark | Mapped task | Coverage | Count | Version | Evidence |
|---|---|---|---|---|---|
| Anthropic Scientific Figure Interpretation Eval root: anthropic-key-life-sciences-evals | Scientific evidence interpretation official-track · high | explicitly-in-scope | Not reported private scientific figure interpretation tasks | reported-2026-01-11 as of 2026-01-11 | anthropic-scientific-figure-evidence |
| BLADE Decision-Discrimination MCQ root: blade | Scientific evidence interpretation official-track · high | explicitly-in-scope | 188 questions individual multiple-choice decision-discrimination questions | arXiv v3 | blade-mcq-evidence-counts |
| LAB-Bench root: lab-bench | Scientific evidence interpretation official-track · high | explicitly-in-scope | Not reported FigQA, LitQA2, SuppQA, and TableQA questions. | repository-998a8e0 | lab-bench-evidence-paper |
| LAB-Bench FigQA root: lab-bench | Scientific evidence interpretation official-track · high | explicitly-in-scope | 226 questions questions across public and private splits | repository-998a8e0 | lab-bench-figqa-evidence-paper |
| LAB-Bench LitQA2 root: lab-bench | Scientific evidence interpretation official-track · high | explicitly-in-scope | 248 questions questions across public and private splits | repository-998a8e0 | lab-bench-litqa2-evidence-paper |
| LAB-Bench SuppQA root: lab-bench | Scientific evidence interpretation official-track · high | explicitly-in-scope | 102 questions questions across public and private splits | repository-998a8e0 | lab-bench-suppqa-evidence-paper |
| LAB-Bench TableQA root: lab-bench | Scientific evidence interpretation official-track · high | explicitly-in-scope | 305 questions questions across public and private splits | repository-998a8e0 | lab-bench-tableqa-evidence-paper |
Official evaluations connected to these benchmarks
Runs are included only for benchmark records mapped here (and formal child tracks when a mapped suite is the root). A task mapping does not imply that every run isolates this task.