LAB-Bench
Benchmark repository for evaluating language models on biology and laboratory research knowledge tasks.
External resource. No execution or independent verification is claimed here.
Can a small non-procedural, rights-reviewed subset support an auditable scoring and contamination-risk review?
Freeze question provenance, answer keys and scoring rules; assess duplicate exposure and avoid treating benchmark success as laboratory competence.
What you could produce
- A version-pinned protocol stating inputs, rights, expected behavior, tolerances and resource limits before execution.
- A retained per-case result and failure ledger with an independently controlled comparison, if qualified execution is later authorized.
Before you use it
- Python/data handling and model backend as selected; benchmark content and optional API costs require separate review.
- A separately qualified isolated runtime with a reviewed, pinned dependency and input closure.
Limits to keep in view
- No source program, example, build hook, package, dataset, model or generated research code has been executed or downloaded as a payload.
- The documented self-contained Python 3.13, 90-second pilot does not establish support for this package, its compiled dependencies, GPUs, services or agent sandboxes.
- Installation success, scientific outcomes, runtime compatibility, latency, memory use, API costs and security properties are unmeasured.
Source and permission context
Preserve the upstream project name, repository, exact source commit and applicable contributor/notices; resolve upstream citation guidance for any later formal use.
Catalog listing reviewed. This review covers the description and source links displayed here.
Reviewed the original LAB-Bench description, repository identity, exact source commit, attribution rule, pinned rights evidence and every unresolved source-rights item. Listing approval concerns the short original description and public source links. The pinned root LICENSE is CC BY-SA 4.0 and its unchanged response digest was refreshed. The pinned README describes an evaluation dataset with a retained private subset. The reviewed short original project description and public repository link copy no questions, figures, protocols, sequences or private content.
Reviewed 2026-09-14. Copying or adapting source files remains subject to their own terms.
repository material designated by the root notice; file-level applicability unresolved · CC-BY-SA-4.0
The complete root LICENSE is the Creative Commons Attribution-ShareAlike 4.0 International legal text. No file-by-file application, third-party figure clearance or payload review has been established. Interpretation is bound to the reviewed response digest at source commit 998a8e0a40cf116c80e1b0e7a805ebb5fb9fa838.
Inspect the license evidence ↗Before copying source material
- Only the cited license and README texts were observed; file exceptions, dependency closure, vendored components and submodules are not comprehensively audited.
- Dataset files, task prompts, generated outputs, model weights, tokenizer assets and hosted APIs are not cleared by a root code license.
- README/documentation reuse rights and version-specific citation guidance remain separately unresolved; only links and original descriptions are retained.
- Question text, figures and third-party scientific source material may have different rights from evaluation code; no biological procedure or dataset payload is copied.