Unverified50% confidenceFactExact time
OpenAI's official website published a benchmark called BioBasics Bench, used to assess a model's ability to understand and analyze everyday biology experiment materials, but the body text explicitly states that only 7 models were tested, and 6.2 Astra was not among them.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
11/18/2026
First Seen
Valid until: 12/2/2026
Sources
Related Entities
Related Claims
UnverifiedBioBasicsBench differs from common bioinformatics or literature question-answering benchmarks in that it targets real operations and judgments at the lab bench.62% similarUnverifiedLeash Bio uses DEL data for model evaluation, which can identify models that score high on standard benchmarks but have poor generalization ability.59% similarUnverifiedThe research team invited domain experts to manually review and annotate 81 cases within BioPhys-Bridge.58% similarUnverifiedThis study is an observational study with a sample size of only 8 domains; it did not conduct significance tests or an intervention experiment of "blocked vs. open" on the same page, so the percentage results should not be extrapolated to other markets or verticals.58% similarUnverifiedBatchBench is a public leaderboard based on the JUMP-CP dataset that specifically evaluates the robustness of biomedical image models on unseen experimental batches and data sources.57% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1185146API
curl https://kongchang.com/api/v1/knowledge/claims/1185146MCP
get_claim(id=1185146)