Research Goal
Design one difficult but fair browsing challenge.
Web research • AI evaluation • Benchmark design
A science question where the discovery, the publication and the prize all carry different dates.
Popular accounts of a scientific result collapse the laboratory work, the paper and the award into a single moment. This benchmark asks only for the date the issue carrying the paper was published, which is recorded by the publisher and can be cross-checked against the two companion papers in the same issue.
Frontier AI browsing systems can answer many questions by locating obvious keywords or matching information from highly ranked webpages. This project explores how to construct a research question that cannot be solved through a single search. The answer must remain short, stable and objectively verifiable, while the clues require the researcher to connect different people, dates, places, organisations, publications, events, records and quantities.
Design one difficult but fair browsing challenge.
A short, stable and objectively verifiable fact.
Every clue must be independently supported by reliable evidence.
A question, answer, clue set, source trail, validation log and JSON record.
AIBRC-008
Portfolio integrity note: Every claim on this page was checked against the cited source before publication. No source, quotation, locator or search result is invented; where a detail could not be verified directly it is described rather than quoted.
Begin with a stable fact that can be confirmed through a reliable primary or authoritative source. Record the exact wording, document title, issuing organisation and publication date.
Develop clues involving different fact types, including dates, people, places, organisations, works, events, records and quantities. Each clue must be independently checkable.
Run obvious, broad and targeted search queries. Record the results to identify accidental shortcuts, direct answer leakage and misleading pages.
Explain how each source supports a specific clue and how the clues logically combine to identify one answer.
Recheck every source, confirm answer uniqueness, document limitations and export the completed benchmark record in a consistent JSON format.
Each clue constrains the answer from a different direction. No single row is sufficient on its own, and every row can be checked against a source another researcher can reach.
Statuses are editable so the matrix can be worked through row by row. Scroll sideways on small screens.
| Clue ID | Fact type | Clue statement | Constraint | Supporting source | Exact locator | Verification status |
|---|---|---|---|---|---|---|
| C01 | Document | The paper is barely one page long and appeared in a weekly British science journal, on pages 737-738 of volume 171. | Fixes the journal, volume and pagination without naming the paper or its date. | Nature — article record for 171:737 | Citation block (volume, pages) | Verified |
| C02 | Person | Its two authors shared a Nobel Prize in Physiology or Medicine nine years later with a third scientist based at a different institution. | Fixes the authors and the 1962 prize, which is a decoy date rather than the answer. | The Nobel Prize — Physiology or Medicine 1962 | Summary page, laureates and citation | Verified |
| C03 | Subject | The structure proposed is a two-chain helix with the chains running in opposite directions around a common axis. | Identifies the content without using the phrase used in every popular retelling. | Nature — article record for 171:737 | Article title and abstract text | Verified |
| C04 | Publication context | Two supporting experimental papers by other groups appear in the same issue, one of them beginning on page 740. | Confirms the issue rather than only the article, so the date asked for is the issue date. | Science History Institute — biography page citations | Citation list: Nature 171 (1953): 737, 738, 740 | Verified |
| C05 | Year | The issue appeared in the spring of 1953. | Narrows the search window without giving the month. | Nature — article record for 171:737 | Publication line | Verified |
Each card records a source that was opened and read before it was cited. Excerpts are quoted only where the wording was read directly; otherwise the record describes what the source shows.
A strong benchmark documents the searches a capable researcher or browsing agent would naturally attempt, including the ones that exposed the answer. Leakage is recorded rather than hidden, and the question is rewritten around it.
| Search number | Exact search query | Search engine or database | Date searched | Top relevant results | Answer revealed? | Problem discovered | Action taken |
|---|---|---|---|---|---|---|---|
| S01 | when was the double helix discovered | 4 September 2026 | Returns the year 1953 and a mixture of February and April anecdotes; no issue date. | No | The obvious query gives a year and a workshop story, not a publication date. | Kept; it shows the question cannot be answered from the popular framing. | |
| S02 | Watson Crick Nature 1953 publication date | 4 September 2026 | Exposes 25 April 1953 in the first results. | Yes | Naming the authors and the journal resolves the task in one step. | The question was rewritten to avoid the author names, the journal name and the words 'double helix', describing the pagination and the prize instead. | |
| S03 | Nature volume 171 pages 737 738 issue date | 4 September 2026 | Reaches the publisher's article record used as E01. | Yes | The pagination is given as a clue, so this path is intended rather than accidental. | Retained: the clue is deliberately a locator, and the solver still has to confirm it is the issue date and not a later online posting. | |
| S04 | Nobel Prize Physiology Medicine 1962 nucleic acids laureates | 4 September 2026 | Returns the awarding body's summary page used as E02. | No | None - it confirms a clue and isolates the decoy year. | Retained as part of the reference path. |
Every candidate source is scored from 1 to 5 against five criteria before it is allowed into the evidence chain.
Who created or issued the information?
Is the source directly connected to the event or record?
Does it support the precise claim being made?
Is the information unlikely to change?
Can another researcher locate the same evidence?
Export a report that maps every clue to the source it rests on, records the verification status and access date of each source, and lists the searches that exposed the answer.
Completion
14 of 15 checks confirmed
93%
The complete challenge is delivered as a single validated JSON document.
aibrc-008-benchmark-record.json
{
"challenge_id": "AIBRC-008",
"title": "Scientific Record Benchmark",
"project_type": "independent_demonstration",
"research_domain": "Science and Publication Records",
"difficulty": "advanced",
"question": "A paper barely a page long, printed on pages 737-738 of volume 171 of a weekly British science journal in the spring of 1953, proposed a two-chain helix whose chains run in opposite directions; its two authors shared a Nobel Prize nine years later. On what date was that issue published?",
"accepted_answers": [
"25 April 1953"
],
"answer_type": "short_text",
"answer_normalisation": {
"case_sensitive": false,
"ignore_punctuation": true,
"accepted_variations": [
"25 April 1953",
"April 25, 1953",
"25/04/1953",
"1953-04-25"
]
},
"clues": [
{
"clue_id": "C01",
"fact_type": "document",
"statement": "The paper is barely one page long and appeared in a weekly British science journal, on pages 737-738 of volume 171.",
"constraint": "Fixes the journal, volume and pagination without naming the paper or its date.",
"source": "Nature — article record for 171:737",
"locator": "Citation block (volume, pages)",
"status": "verified"
},
{
"clue_id": "C02",
"fact_type": "person",
"statement": "Its two authors shared a Nobel Prize in Physiology or Medicine nine years later with a third scientist based at a different institution.",
"constraint": "Fixes the authors and the 1962 prize, which is a decoy date rather than the answer.",
"source": "The Nobel Prize — Physiology or Medicine 1962",
"locator": "Summary page, laureates and citation",
"status": "verified"
},
{
"clue_id": "C03",
"fact_type": "subject",
"statement": "The structure proposed is a two-chain helix with the chains running in opposite directions around a common axis.",
"constraint": "Identifies the content without using the phrase used in every popular retelling.",
"source": "Nature — article record for 171:737",
"locator": "Article title and abstract text",
"status": "verified"
},
{
"clue_id": "C04",
"fact_type": "publication context",
"statement": "Two supporting experimental papers by other groups appear in the same issue, one of them beginning on page 740.",
"constraint": "Confirms the issue rather than only the article, so the date asked for is the issue date.",
"source": "Science History Institute — biography page citations",
"locator": "Citation list: Nature 171 (1953): 737, 738, 740",
"status": "verified"
},
{
"clue_id": "C05",
"fact_type": "year",
"statement": "The issue appeared in the spring of 1953.",
"constraint": "Narrows the search window without giving the month.",
"source": "Nature — article record for 171:737",
"locator": "Publication line",
"status": "verified"
}
],
"evidence": [
{
"evidence_id": "E01",
"claim_supported": "The answer, and clues C01, C03 and C05.",
"source_title": "Molecular Structure of Nucleic Acids: A Structure for Deoxyribose Nucleic Acid",
"issuing_organisation": "Nature (Springer Nature)",
"source_type": "Peer-reviewed journal record",
"url": "https://www.nature.com/articles/171737a0",
"publication_date": "25 April 1953",
"page_number": "737-738",
"section": "Article record header: volume, pages and publication date",
"access_date": "4 September 2026",
"reliability": "High - the publisher of record",
"verification_status": "verified"
},
{
"evidence_id": "E02",
"claim_supported": "Clue C02 and the neutralisation of the 1962 decoy.",
"source_title": "The Nobel Prize in Physiology or Medicine 1962",
"issuing_organisation": "The Nobel Foundation",
"source_type": "Awarding institution",
"url": "https://www.nobelprize.org/prizes/medicine/1962/summary/",
"publication_date": "Continuously maintained page",
"page_number": "n/a (web page)",
"section": "Prize summary and citation",
"access_date": "4 September 2026",
"reliability": "High - the awarding body",
"verification_status": "verified"
},
{
"evidence_id": "E03",
"claim_supported": "Clue C04 - independent confirmation of the issue contents.",
"source_title": "Francis Crick, Rosalind Franklin, James Watson, and Maurice Wilkins",
"issuing_organisation": "Science History Institute",
"source_type": "Museum and research institute",
"url": "https://www.sciencehistory.org/education/scientific-biographies/james-watson-francis-crick-maurice-wilkins-and-rosalind-franklin/",
"publication_date": "Continuously maintained page",
"page_number": "n/a (web page)",
"section": "Citations to Nature volume 171",
"access_date": "4 September 2026",
"reliability": "Medium-high - a specialist institution, used only for corroboration",
"verification_status": "verified"
}
],
"search_validation": [
{
"search_id": "S01",
"query": "when was the double helix discovered",
"platform": "Google",
"search_date": "4 September 2026",
"relevant_results": "Returns the year 1953 and a mixture of February and April anecdotes; no issue date.",
"answer_exposed": false,
"notes": "The obvious query gives a year and a workshop story, not a publication date. Kept; it shows the question cannot be answered from the popular framing."
},
{
"search_id": "S02",
"query": "Watson Crick Nature 1953 publication date",
"platform": "Google",
"search_date": "4 September 2026",
"relevant_results": "Exposes 25 April 1953 in the first results.",
"answer_exposed": true,
"notes": "Naming the authors and the journal resolves the task in one step. The question was rewritten to avoid the author names, the journal name and the words 'double helix', describing the pagination and the prize instead."
},
{
"search_id": "S03",
"query": "Nature volume 171 pages 737 738 issue date",
"platform": "Google",
"search_date": "4 September 2026",
"relevant_results": "Reaches the publisher's article record used as E01.",
"answer_exposed": true,
"notes": "The pagination is given as a clue, so this path is intended rather than accidental. Retained: the clue is deliberately a locator, and the solver still has to confirm it is the issue date and not a later online posting."
},
{
"search_id": "S04",
"query": "Nobel Prize Physiology Medicine 1962 nucleic acids laureates",
"platform": "Google",
"search_date": "4 September 2026",
"relevant_results": "Returns the awarding body's summary page used as E02.",
"answer_exposed": false,
"notes": "None - it confirms a clue and isolates the decoy year. Retained as part of the reference path."
}
],
"quality_checks": {
"The answer is short and objectively verifiable.": true,
"The answer is supported by an authoritative source.": true,
"The question has only one defensible answer.": true,
"No single clue directly exposes the answer.": true,
"Each clue is independently checkable.": true,
"The clue set spans multiple fact types.": true,
"The task requires multi-step reasoning.": true,
"Obvious searches have been documented.": true,
"Search-result snippets do not reveal the answer.": false,
"Sources are publicly accessible.": true,
"Exact pages, tables or sections are recorded.": true,
"The evidence chain can be reproduced.": true,
"The question avoids subjective interpretation.": true,
"The answer is stable over time.": true,
"No fabricated evidence is included.": true
},
"researcher": {
"name": "TMK Research Lab",
"portfolio": "https://theophilusmorkporkpor.lovable.app",
"email": "tkudjordjie@gmail.com"
}
}AI-generated statements are never treated as evidence. Every factual claim must be verified against an accessible external source.
This benchmark is complete: the answer is fixed, 3 evidence records are attached to specific clues, and 4 validation searches are documented, including those that exposed the answer. It demonstrates the ability to approach an unfamiliar subject systematically, separate evidence from assumption and document research so that another person can audit it.
Three dates compete: the laboratory work in the winter of 1953, the issue date in April, and the prize in 1962. Only one is a published record.
The plain discovery query returned a year and an anecdote, and never a publication date.
The publisher's article record, because it carries the issue date as a structured field next to the volume and pagination.
The question asks for the issue date and gives the exact pages, so a solver cannot substitute the discovery month or the award year.
Answering 1962, answering 28 February 1953 from the laboratory anecdote, or giving a modern online-posting date from a reprint page.
The pagination clue is a strong locator, so a solver who searches it directly reaches the answer quickly.
Require the first page of the third paper in the same issue, which forces the solver to open the issue contents rather than the single article record.
I am available for web research, fact-checking, AI evaluation, benchmark construction, source verification and structured research documentation.
TMK Research Labtkudjordjie@gmail.comhttps://theophilusmorkporkpor.lovable.apphttps://www.linkedin.com/in/theophilus-kudzordzi