1,000 works · Opus 4.8 · GPT-5.6 (high) · Grok 4.5 · one locked rubric
The field’s boundary is not a line. It is a region.
Three frontier models screened the same 1,000 works, drawn from the real frame with known selection probabilities, against the same locked rubric on its full eight-field payload. Of the 51 works any model called metaresearch, only 19 (37.3%) were called metaresearch by all three, and 24 (47.1%) rest on a single model's opinion.
Two screeners can agree on a rate while finding almost entirely different works. At a ~1% base rate, the settled rejects buy 98% agreement for free — which is why an agreement statistic computed over the whole sample tells you nothing about the boundary, and why the table below, not a percentage, is the deliverable.
The three models are not interchangeable
How many of the 1,000 works each model called metaresearch (tier T1 or T2), on identical input. T3 is adjacent and does not count as in scope, which is why a model's count can never exceed the 51 works in the dossier.
Swap which model you call “the screener” and the size of the field moves. That spread — not the binomial confidence interval on any one model's labels — is the honest uncertainty on how big Canadian metaresearch is.
The disagreement dossier
All three verdicts, confidences and reasons side by side. These are the works against which the inclusion criteria have to be written — because these are the works on which reasonable screeners, given the same rubric and the same evidence, disagree.
| Work | Stratum | n_in | Opus 4.8 | GPT-5.6 | Grok 4.5 |
|---|---|---|---|---|---|
| Influence of Nitrogen Sources or Plant Growth Regulator on Improving Spring Canola Yield in Northwestern Ontario 2023 · article · en · Journal of Agricultural Science and Technology B | about_only w=3321 | 0/3 | OUThigh empirical Agronomy trial on nitrogen sources and canola yield; abstract absent but the title is unambiguous. | OUTlow empirical The title clearly concerns canola yield in agriculture, not research itself, but the abstract is missing. | OUTmedium empirical Agronomy of canola yield under nitrogen or growth regulators; title is domain agriculture with no abstract. |
| Systematic pole-zero sorting method for neuro-TF modeling of electromagnetic response 2024 · article · en · Optics Express | aff_core w=5595 | 0/3 | OUThigh empirical Engineering paper on a pole-zero sorting method for electromagnetic response modeling; the object is a modeling technique in electronics. | OUThigh empirical The paper develops a modeling method for electromagnetic filter responses, not research practice. | OUThigh empirical Engineering parametric modeling of EM filter responses, not research about research. |
| Armand Mattelart : repères biographiques 2015 · book-chapter · fr · HAL (Le Centre pour la Communication Scientifique Directe) | french w=1554 | 0/3 | T3low other Biographical chronology of a communication scholar; contextual material about a researcher's career, akin to the obituary/biographical items the rubric places in T3, and the record has no real abstract. | OUThigh other It is a biographical account of a scholar, not a study of research itself. | OUTlow other Biographical notes on Armand Mattelart; thin abstract and not about research as object. |
| Chapter 8 Canada, Quebec, and La Francophonie 2018 · book-chapter · en · University of Ottawa Press eBooks | about_only w=3321 | 0/3 | OUTmedium other Book chapter on Canada, Quebec and la Francophonie; political relations, not the research system. | OUTlow conceptual The title identifies a chapter about Canada, Quebec, and Francophonie but provides too little information to indicate metaresearch. | OUTlow other Chapter on Canada, Quebec, and La Francophonie with empty abstract; no evidence object is research rather than politics/culture. |
| Cahier couleur 2016 · book-chapter · fr · Presses universitaires de Rennes eBooks | french w=1554 | 0/3 | OUThigh other Book insert listing colour illustrations; no research content. | OUThigh other The chapter consists of historical and geographic illustrations rather than metaresearch. | OUThigh other Color-plate figure insert for a book on borders, not research as object. |
| Les milieux alternatifs de scolarisation : de nouveaux entrepreneurs sociaux ? 2024 · article · fr · Les Sciences de l éducation - Pour l Ère nouvelle | french w=1554 | 0/3 | OUThigh empirical Case studies of alternative schooling settings as social entrepreneurship; the object is education provision. | OUThigh empirical The study examines social entrepreneurship and alternative schooling in Quebec. | OUThigh empirical Social entrepreneurship framing of alternative Quebec schooling; education/society, not research systems. |
| Abrupt Monetary Policy Change and Unanchoring of Inflation Expectations 2024 · article · en · SSRN Electronic Journal | aff_core w=5595 | 0/3 | OUTmedium empirical Economics paper on monetary policy and inflation expectations with no abstract; a macroeconomics question. | OUTlow empirical The title indicates an economics study rather than metaresearch, but the abstract is missing. | OUTlow other Title indicates monetary-policy economics; no abstract, so judged title-only and clearly not metaresearch. |
| Freedom of religion : Comparative perspectives France-Canada 2015 · preprint · fr | french w=1554 | 0/3 | OUThigh conceptual Comparative legal analysis of freedom of religion in France and Canada; constitutional law, not the research system. | OUThigh conceptual The dissertation compares constitutional approaches to religious freedom rather than studying research. | OUThigh conceptual Comparative constitutional law on religious freedom in France and Canada, not the research system. |
| Spectrotemporal cues and attention jointly modulate fMRI network topology for sentence and melody perception 2024 · article · en · Scientific Reports | aff_core w=5595 | 0/3 | OUThigh empirical fMRI study of speech and melody perception and network topology. | OUThigh empirical It studies neural processing of speech and music using fMRI. | OUThigh empirical fMRI network study of speech and music perception is cognitive neuroscience. |
| Importance of Product-Specific Testing in Determining Durability Reduction Factor for Polyester Geogrids in High pH Conditions 2024 · article · en | aff_core w=5595 | 0/3 | OUThigh empirical Geotechnical testing of polyester geogrid durability; the standards at issue are engineering design standards, not research standards. | OUThigh empirical It experimentally evaluates geogrid durability under alkaline conditions. | OUThigh empirical Geogrid durability product testing is civil/geotechnical engineering. |
| The risk factors for arteriovenous fistula dysfunction in maintenance hemodialysis patients: A cross‐sectional study 2024 · article · en · Hemodialysis International | venue_new w=2684 | 0/3 | OUThigh empirical Cross-sectional clinical study of risk factors for arteriovenous fistula dysfunction in hemodialysis patients. | OUThigh empirical It studies clinical risk factors for arteriovenous-fistula dysfunction, not research practice. | OUThigh empirical Clinical risk factors for AVF dysfunction in dialysis; medical object. |
| Adult Services and Employment for Adults with Autism: Options, Challenges, and Hope 2024 · book-chapter · en | aff_core w=5595 | 0/3 | OUThigh other Book chapter on adult services and employment for adults with autism; the object is disability service provision. | OUTlow conceptual The title concerns employment and services for autistic adults, not research itself. | OUTlow other Book chapter on autism adult services/employment; empty abstract, domain social/clinical topic. |
| Aparté sur le sens de la douleur 2023 · article · fr · Petite revue de philosophie | french w=1554 | 0/3 | OUTmedium conceptual French-language philosophical essay on the meaning of pain; the object is suffering, not research. | OUThigh conceptual It is a philosophical discussion of pain, not research itself. | OUTmedium editorial/commentary Brief philosophical aside on the meaning of pain; not about research practice. |
| Review of the subgenera Metopius Panzer and Peltastes Illiger of the genus Metopius Panzer (Hymenoptera: Ichneumonidae: Metopiinae) from Japan 2024 · article · en · Zootaxa | aff_core w=5595 | 0/3 | OUThigh empirical Taxonomic revision of Metopius wasp subgenera in Japan; insect systematics. | OUThigh empirical It revises the taxonomy of Japanese wasps, not research practice. | OUThigh empirical Taxonomic review of Japanese Metopius wasps; entomology systematics. |
| De la ouananiche (Salmo salar) dans une rivière en milieu agricole près de chez vous : le cas de la Petite rivière Eusèbe au Saguenay–Lac-Saint-Jean 2024 · article · fr · Le Naturaliste canadien | french w=1554 | 0/3 | OUThigh empirical Freshwater ecology note on landlocked salmon in an agricultural watershed. | OUThigh empirical The article concerns landlocked salmon in a Quebec river, not research practice. | OUTmedium other Natural-history note on landlocked salmon in a Quebec agricultural river; domain ecology. |
| The frequency of metal enrichment of cool helium-atmosphere white dwarfs using the DESI early data release 2024 · article · en · Monthly Notices of the Royal Astronomical Society Letters | aff_core w=5595 | 0/3 | OUThigh empirical Astronomy paper on the frequency of metal enrichment in helium-atmosphere white dwarfs using DESI data. | OUThigh empirical It estimates planetary material frequency around white dwarfs, not research practice. | OUThigh empirical Astrophysics measurement of metal enrichment frequency in cool helium white dwarfs. |
| Gecko-inspired dry adhesives for heritage conservation – tackling the surface roughness with empirical testing and finite element modelling 2023 · dataset · en · Figshare | aff_core w=5595 | 0/3 | OUThigh empirical Materials testing of gecko-inspired dry adhesives for heritage conservation. | OUThigh empirical The study tests adhesives for heritage conservation rather than research methods or systems. | OUThigh empirical Materials testing of gecko-inspired adhesives for heritage conservation, domain applied science. |
| The clinical effects of individual career counseling on clients’ psychological distress 2024 · article · en · The Career Development Quarterly | aff_core w=5595 | 0/3 | OUThigh empirical Study of the clinical effects of career counseling on psychological distress; the object is counseling outcomes. | OUThigh empirical The study evaluates clinical effects of career counseling, not research itself. | OUThigh empirical Clinical outcomes of career counseling for client distress; counseling effectiveness, not research systems. |
| Challenges and opportunities in achieving secure hospital clinical mobility management: An illustrative use case 2024 · article · en · Journal of Hospital Administration | venue_new w=2684 | 0/3 | OUThigh empirical Health IT use case on managing shared clinical mobile devices in a hospital system. | OUThigh empirical It describes a hospital technology-management use case, not research practice. | OUThigh empirical Hospital shared mobile-device management use case is clinical IT operations, not metaresearch. |
| Balancing of Robustness and Performance for Triple Inverted Pendulum Using μ-Synthesis and Gazelle Optimization 2024 · article · en · Journal Européen des Systèmes Automatisés | venue_new w=2684 | 0/3 | OUThigh empirical Robust control design for a triple inverted pendulum using mu-synthesis and a metaheuristic; a control engineering contribution. | OUThigh empirical The paper develops and simulates a control strategy for an inverted pendulum, not research practice. | OUThigh empirical Control-engineering paper balancing robustness and performance of a triple inverted pendulum. |
| Device-specific quality of life: results from the ATLAS trial—avoid transvenous leads in appropriate subjects 2024 · article · en · European Journal of Cardiovascular Nursing | aff_core w=5595 | 0/3 | OUThigh empirical Patient-reported quality of life outcomes from a defibrillator trial; a clinical result, not a study of trial practice. | OUThigh empirical This is a clinical trial outcomes analysis, and methodological terms are incidental. | OUThigh empirical Clinical trial report of ICD device quality of life; object is medical device outcomes, not research methods. |
| Development of an mRNA-based therapeutic vaccine mHTV-03E2 for high-risk HPV-related malignancies 2024 · article · en · Molecular Therapy | aff_core w=5595 | 0/3 | OUThigh empirical Development of an mRNA therapeutic vaccine for HPV-related malignancies; the object is a vaccine. | OUThigh empirical It concerns development of a therapeutic cancer vaccine, not research practice. | OUThigh other Therapeutic HPV mRNA vaccine development; biomedical product research, not metaresearch. |
| Comparison of three diagnostic strategies for suspicion of pulmonary embolism: planar ventilation-perfusion scan (V/Q), CT pulmonary angiography (CTPA) and single photon emission CT ventilation-perfusion scan (SPECT V/Q): a protocol of a randomised controlled trial 2024 · article · en · BMJ Open | aff_core w=5595 | 0/3 | OUThigh empirical Protocol for a randomized trial comparing imaging strategies for diagnosing pulmonary embolism; the strategies are clinical diagnostic methods, not research methods. | OUThigh empirical The protocol evaluates diagnostic strategies for pulmonary embolism. | OUThigh empirical RCT protocol comparing PE diagnostic imaging strategies; clinical diagnostic research. |
| Long-Term Safety and Efficacy of Voclosporin in Patients With Lupus Nephritis and Low eGFR 2022 · article · en · Journal of the American Society of Nephrology | aff_core w=5595 | 0/3 | OUThigh empirical Post-hoc analysis of voclosporin safety and efficacy in lupus nephritis with low eGFR; a clinical trial analysis. | OUThigh empirical The analysis evaluates the safety and efficacy of a lupus nephritis treatment, not research itself. | OUThigh empirical Clinical efficacy and safety of voclosporin in lupus nephritis; therapeutic domain study. |
| Tegoprubart for the Prevention of Rejection in Kidney Transplant: Update of Emerging Data from an Ongoing Trial 2023 · article · en · Journal of the American Society of Nephrology | aff_core w=5595 | 0/3 | OUThigh empirical Interim clinical trial results for an antibody preventing kidney transplant rejection. | OUThigh empirical The work reports a kidney-transplant treatment trial rather than studying research methods. | OUThigh empirical Clinical transplant trial update on an immunosuppressive drug, domain medicine. |
| Lifespan Effects of Muscle-specific Androgen Receptor Overexpression on Body Composition of Male and Female Rats 2024 · article · en · Physiology | aff_core w=5595 | 0/3 | OUThigh empirical Physiology experiment on muscle androgen receptor overexpression and body composition in rats. | OUThigh empirical It studies androgen receptor effects in rats, not research practice. | OUThigh empirical Physiology study of androgen receptor overexpression and body composition in rats. |
| Les changements climatiques et les critères et indicateurs d’aménagement durable des forêts : plaidoyer pour une vision élargie des relations communauté-forêt au Canada 2024 · article · fr · VertigO | french w=1554 | 0/3 | OUTmedium conceptual Argues for revising criteria and indicators of sustainable forest management under climate change; forest policy indicators, not research evaluation. | OUThigh conceptual This article discusses forest-management indicators and climate adaptation, not indicators of research. | OUThigh conceptual Forest-community sustainability indicators under climate change in Canada; forest policy, not the Canadian research system. |
| Implementing intersectoral alcohol policies at the local level: A case study from Santiago, Chile, 2014-2017 2024 · article · en · The International Journal of Alcohol and Drug Research | venue_new w=2684 | 0/3 | OUThigh empirical Case study of municipal alcohol control policy implementation; health policy, not research policy. | OUThigh empirical The work evaluates local alcohol-control policy implementation in Santiago. | OUThigh empirical Case study of local alcohol-control policy implementation in Chile; public health policy, not research practice. |
| Table des Matières 2022 · book-chapter · fr · Presses de l'Université du Québec eBooks | french w=1554 | 0/3 | OUThigh other Book front matter (table of contents); no research content. | OUTlow other The title indicates a table of contents with no substantive information about the object of study. | OUTlow other Table of contents with no abstract; not a study of research practice. |
| Epinephrine Dosing Intervals Are Associated With Pediatric In-Hospital Cardiac Arrest Outcomes: A Multicenter Study* 2024 · article · en · Critical Care Medicine | fund_new w=1679 | 0/3 | OUThigh empirical Secondary analysis of trial data on epinephrine dosing intervals and pediatric cardiac arrest outcomes; a clinical question. | OUThigh empirical It evaluates epinephrine dosing intervals and cardiac-arrest outcomes, not research practice. | OUThigh empirical Pediatric cardiac-arrest outcomes vs epinephrine dosing intervals; clinical resuscitation research. |
| Degrees are Useless in SNORT When Measuring Temperature 2024 · preprint · en · arXiv (Cornell University) | fund_new w=1679 | 0/3 | OUThigh conceptual Combinatorial game theory result on temperature in the game SNORT; pure mathematics. | OUThigh conceptual This mathematical paper studies temperatures in a combinatorial game, not research. | OUThigh conceptual Combinatorial game theory analysis of Snort temperature; not measurement science of research. |
| Progress of plant‐derived non‐starch polysaccharides and their challenges and applications in future foods 2024 · review · en · Comprehensive Reviews in Food Science and Food Safety | aff_core w=5595 | 0/3 | OUThigh conceptual Review of plant-derived polysaccharides for future foods; food science. | OUThigh other This review concerns plant polysaccharides and food applications rather than research itself. | OUThigh other Domain review of plant non-starch polysaccharides for future foods; food science, not metaresearch. |
| VERS UN CHEMIN DE VIE EN VÉRITÉ 2021 · article · fr · Science et Esprit | french w=1554 | 0/3 | OUThigh conceptual Theological essay on truth, faith and existence; not about research. | OUThigh conceptual The article develops a theological account of truth and faith rather than studying research. | OUThigh conceptual Theological-philosophical essay on truth and faith; not about research as object. |
| eLife Assessment: Antigenic drift and subtype interference shape A(H3N2) epidemic dynamics in the United States 2024 · peer-review · en | aff_core w=5595 | 0/3 | OUThigh other eLife peer-review assessment of an influenza epidemiology paper; it is a peer review artifact whose object is viral dynamics, not a study of peer review. | OUThigh editorial/commentary The work studies influenza epidemic dynamics rather than research practice. | OUThigh other eLife assessment of an influenza epidemiology paper; peer review as document, not study of peer review or research practice. |
| Effects of forage feeding level on ruminal pH and metabolic adaptation of the rumen epithelium in pre-weaned Jersey calves 2024 · article · en · Veterinary and Animal Science | aff_core w=5595 | 0/3 | OUThigh empirical Animal nutrition trial on forage feeding in calves; domain science. | OUThigh empirical This experiment studies calf nutrition and rumen physiology, not research. | OUThigh empirical Animal nutrition physiology study of calf rumen pH and SCFA transport. |
| TOI-757 b: an eccentric transiting mini-Neptune on a 17.5-d orbit 2024 · preprint · en · arXiv (Cornell University) | fund_new w=1679 | 0/3 | OUThigh empirical Spectroscopic characterization of an exoplanet; astronomy. | OUThigh empirical This characterizes an exoplanet and its formation, not research practice. | OUThigh empirical Exoplanet discovery and characterization of TOI-757 b; domain astronomy. |
| Taxonomy of Leodice Lamarck, 1818 (Annelida, Eunicidae) from Northeast Brazil (Tropical Southwest Atlantic Ocean) with description of a new species 2024 · article · en · Zootaxa | aff_core w=5595 | 0/3 | OUThigh empirical Polychaete taxonomy with description of a new species. | OUThigh empirical The work presents a taxonomic study of marine worms, not a study of research. | OUThigh empirical Taxonomic description of marine annelids; object is biodiversity systematics. |
| De l’acadianité en contexte louisianais : <i>Les Veillées d’une soeur ou le destin d’un brin de mousse</i> (1877), autobiographie de Désirée Martin 2020 · book-chapter · fr | french w=1554 | 0/3 | OUThigh conceptual Literary studies chapter on Acadian identity in a nineteenth-century Louisiana autobiography; no research-practice object. | OUThigh conceptual This book chapter analyzes a literary autobiography and does not study research practice. | OUThigh conceptual Literary/historical chapter on Acadian identity in a Louisiana autobiography. |
| Resilience Informatics: Role of Informatics in Enabling and Promoting Public Health Resilience to Pandemics, Climate Change, and Other Stressors 2024 · article · en · Interactive Journal of Medical Research | venue_new w=2684 | 0/3 | OUThigh editorial/commentary Viewpoint on informatics for public health resilience; commentary on health technology, not on research practice. | OUThigh conceptual This viewpoint concerns public-health resilience technologies and does not study research itself. | OUThigh conceptual Viewpoint on informatics for public-health resilience to pandemics/climate; applied public health, not studying research. |
| The 3D Lyman-<i>α</i> forest power spectrum from eBOSS DR16 2024 · article · en · Monthly Notices of the Royal Astronomical Society | fund_new w=1679 | 0/3 | OUThigh empirical Cosmology measurement of the 3D Lyman-alpha forest power spectrum from eBOSS; the object is large-scale structure. | OUThigh empirical The paper measures a cosmological power spectrum and evaluates its modeling pipeline. | OUThigh empirical Cosmology measurement of the 3D Lyman-alpha forest power spectrum. |
| Editorial: Experience with quantum annealing computation 2024 · editorial · en · Frontiers in Computer Science | aff_core w=5595 | 0/3 | OUTmedium editorial/commentary No abstract; editorial introducing a research-topic collection on quantum annealing, i.e. commentary about a domain literature rather than about research practice. | OUThigh editorial/commentary The editorial concerns quantum annealing computation as a scientific topic, not research practice. | OUTmedium editorial/commentary Editorial on quantum annealing computation experience; domain computing topic, not research systems (no abstract). |
| Towards the Enhancement of Buildings’ Sustainability: IoT-Based Building Management Systems (IoT-BMS) 2024 · article · en · IOP Conference Series Earth and Environmental Science | aff_core w=5595 | 0/3 | OUThigh empirical IoT-based building management system for energy efficiency; an engineering solution. | OUThigh empirical This work presents an IoT building-management solution and does not study research itself. | OUThigh empirical IoT building-management systems for energy efficiency; applied engineering, not a study of research. |
| FPGA based Time-To-Digital Converter for Time of Flight Computed Tomography 2024 · article · en | aff_core w=5595 | 0/3 | OUThigh empirical FPGA time-to-digital converter for time-of-flight CT; instrumentation engineering. | OUThigh empirical This is an engineering study of a time-to-digital converter for medical imaging, not of research practice. | OUThigh empirical Domain medical-imaging hardware engineering; object is a TDC device, not research practice. |
| Empathy and High-Fidelity Human Patient Simulators: A Critical Analysis of Undergraduate Nursing Education 2024 · article · en · Quality Advancement in Nursing Education - Avancées en formation infirmière | venue_new w=2684 | 0/3 | OUTmedium conceptual Discussion paper critiquing high-fidelity simulation for empathy development in undergraduate nursing education; the object is pedagogy and professional training for practice, not research practice. | OUThigh conceptual This discussion concerns simulation-based nursing education and empathy, not research itself. | OUThigh conceptual Critique of simulators in undergraduate nursing education for empathy; professional pedagogy, not research-as-object. |
| Redefining medical education: harnessing the power of patient feedback 2024 · article · en · Frontiers in Medicine | about_only w=3321 | 0/3 | OUThigh editorial/commentary Viewpoint advocating patient feedback in clinical training; the object is medical education, not the research workforce or research practice. | OUThigh editorial/commentary This viewpoint advocates patient feedback in medical education, not the conduct of research. | OUThigh editorial/commentary Viewpoint on patient feedback in clinical medical education, not research training or metaresearch. |
| Internalizing ASR with Implicit Chain of Thought for Efficient Speech-to-Speech Conversational LLM 2024 · preprint · en · arXiv (Cornell University) | fund_new w=1679 | 0/3 | OUThigh empirical Machine learning method internalizing ASR chain of thought in a speech-to-speech conversational LLM; a model architecture contribution. | OUThigh empirical The preprint develops a speech-to-speech language-model method, not a study of research practice. | OUThigh empirical Speech LLM engineering for ASR chain-of-thought; AI systems research, not metaresearch. |
| Which failures do patient‐specific quality assurance systems need to catch? 2024 · article · en · Medical Physics | aff_core w=5595 | 0/3 | OUThigh empirical Failure-mode analysis for radiotherapy patient-specific quality assurance; 'quality assurance' here is clinical dosimetry, not research practice (polysemy). | OUThigh empirical This study evaluates clinical radiotherapy quality assurance rather than research quality or practice. | OUThigh empirical Clinical radiotherapy patient-specific QA failure modes; quality-assurance polysemy, not research integrity or methods research. |
| Gérard Noiriel, Préférence nationale. Leçon d’histoire à l’usage des contemporains 2024 · article · fr · Hommes & migrations | french w=1554 | 0/3 | OUThigh editorial/commentary Book review of a historian's essay on national preference and immigration politics; domain commentary. | OUThigh other This is a book review about immigration history and does not study research practice. | OUThigh editorial/commentary Book review of Noiriel on national-preference immigration history; not about research practice. |
| PATH-34. IT NEVER RAINS BUT IT POURS - DIAGNOSING ANAPLASTIC PLEOMORPHIC XANTHASTROCYTOMA TWICE IN A FORTNIGHT - OUR EXPERIENCE IN A RESOURCE CONSTRAINED SETTING 2024 · article · en · Neuro-Oncology | aff_core w=5595 | 0/3 | OUThigh empirical Neuro-oncology case report of two brain tumor diagnoses. | OUThigh empirical This reports clinical case experiences involving a rare brain tumor, not research practice. | OUThigh empirical Clinical case reports of rare brain tumours and their diagnosis/treatment. |
| A β-hydroxybutyrate shunt pathway generates anti-obesity ketone metabolites 2024 · article · en · Cell | aff_core w=5595 | 0/3 | OUThigh empirical Biochemistry of a ketone body shunt pathway and anti-obesity metabolites. | OUThigh empirical This is a biomedical study of ketone metabolism and obesity in mice and humans. | OUThigh empirical Biochemistry paper identifying a ketone metabolite pathway; domain molecular biology. |
What the screen actually found
Three frontier models (Opus 4.8, GPT-5.6 high, Grok 4.5) screened the same 1,000 works from the REAL 4.3M frame, on the rubric's FULL eight-field payload, with randomized manifest-logged chunks and harness-written labels: every defect D1, D2, D11 and finding 16 identified, repaired. Design-weighted base rates span 1.89% to 3.54% (1.9x). But the sets are the finding, as finding 16 predicted: of the 51 works ANY model called metaresearch, only 19 (37%) were called metaresearch by ALL THREE, and 24 (47%) rest on a single model's opinion. THE FIELD'S BOUNDARY IS NOT A LINE THE MODELS SHARE; IT IS A REGION THEY EACH CUT DIFFERENTLY. GPT-5.6 also violated the locked output schema on 18 of 1,000 records, writing genre values into the tier field, which the manifest validator caught. The deliverable is not the base rate: it is the disagreement dossier, the 51 works that mark the empirical boundary of the field and against which the inclusion criteria must actually be written.
Computed 2026-07-12T19:52:14Z · all 22 findings · this dossier as JSON