Useful science.
Honest context.
A short explanation should make research easier to understand without making it sound more certain.
What is in the library?
This edition contains 17,820 publication records across 14 topics, including 2,531 short explanations. The original collection was deduplicated by PubMed ID and checked against PubMed. Older tips, automatically extracted sample sizes and old DOI links were not treated as verified scientific findings.
Three labels, different levels of detail
Reviewed explanation: a short explanation with linked primary sources and the research context checked during the initial editorial review. Abstract-based explanation: an explanation drafted from the current published abstract and checked by a separate AI critic; the full paper has not been reviewed. Paper record: source details have been checked, but no explanation has passed editorial review yet.
There are currently 2,523 abstract-based explanations. Records with unresolved sources or publication alerts are labeled separately. A review label does not mean the evidence is conclusive.
Keep the result and its limits together
Each explanation separates the study’s question, design, participants and findings from interpretation. Numbers retain their comparison, time frame and units. An association does not establish a cause; a short trial does not establish a lifelong benefit; a questionnaire score does not show clinical recovery.
How the collection was checked
Publication titles, identifiers, study types and linked notices were refreshed from PubMed. When a DOI could not be verified it was omitted. Retractions and expressions of concern are displayed as publication alerts. A checked record is not a full-text appraisal or a systematic review of all available evidence.
Who does the work?
100HP editorial describes the AI-assisted research, writing and development process. Separate AI critics review source fidelity, clarity, usability, visual consistency and technical quality. There is no claim of an independent human scientific or clinical review. Abstract-based drafts must pass both structural checks and a separate source-fidelity review before they appear as explanations.
A suggestion is not an experimental result
“One idea to consider” is labeled editorial interpretation. It does not prescribe personal treatment or guarantee an outcome. Read the evidence in its context before deciding what is useful to you.
Make room for corrections
Each record has a stable PubMed identifier. Source check dates and explanation review dates describe different steps. Linked publication notices appear alongside the source. This edition does not continuously monitor new research or replace professional care.