We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
PROMETHEUS: Automating Deep Causal Research Integrating Text, Data and Models
AI summary Read the abstract
Researchers built an AI tool called PROMETHEUS that reads through scientific studies, data, and code to map out cause-and-effect claims—like whether GLP-1 drugs actually cause weight loss, or whether resveratrol in red wine really benefits health—and flags where the evidence is strong, weak, or contradictory. This matters because health claims in the news often oversimplify messy science; this tool could help researchers (and eventually the public) see when "proven" health effects are actually shaky or based on conflicting studies. Notably, one test case applied this approach to a climate study on microplastics, showing the tool can also stress-
Large language models can extract local causal claims from text, but those claims become more useful when organized as persistent, navigable world models rather than as flat summaries. We introduce PROMETHEUS, a framework that turns retrieved literature, filings, reviews, reports, agent traces, source data, code, simulations, and scientific models into causal atlases: sheaf-like families of local causal predictive-state models over an explicit cover of a research substrate. Each local region contains causal episodes, structured claim tables, predictive tests, support statistics, and provenance; restriction maps compare overlapping regions; gluing diagnostics expose agreement, drift, contradiction, and underdetermination. The resulting Topos World Model is not a single universal graph. It is a research instrument for navigating what a corpus says, where it says it, how strongly it is supported, and where local claims fail to assemble into a coherent global view. Three literature-atlas case studies -- ocean-temperature impacts on marine populations, GLP-1 weight-loss evidence, and resveratrol/red-wine health-benefit claims -- illustrate deep causal research from text with explicit locality, evidence, persistent state, and gluing tension. Four grounded-counterfactual case studies -- a Nature Climate Change microplastics forcing paper, an Indus Valley hydrology paper with VIC-derived figure data and model code, the canonical Sachs protein-signaling study with single-cell perturbation data, and a Nature singing-mouse study with MAPseq projection matrices -- show a stronger mode: when a paper ships source data, simulation outputs, or code, PROMETHEUS can evaluate a counterfactual against that scientific substrate and then rebuild the sheaf world model around the
More Papers Like This
PROMETHEUS: Automating Deep Causal Research Integrating Text, Data and Models
AI summary Read the abstract
Researchers built an AI tool called PROMETHEUS that reads through scientific papers, data, and even computer code to map out cause-and-effect claims—showing not just what studies say, but how strong the evidence is and where different studies actually contradict each other. This matters for health-conscious readers because the tool was tested on real controversial topics like GLP-1 weight-loss drugs and resveratrol (the "red wine health benefit" compound), helping reveal when popular health claims are well-supported versus shaky or overhyped. It was also tested on a microplastics climate-impact study, showing the tool can even check claims against the raw under
Ensuring reliable feature importance in food chemistry AI
AI summary Read the abstract
Researchers demonstrated that standard machine learning models applied to a microplastic-cancer dataset can produce misleading feature importance scores because they optimize prediction rather than causal inference, and proposed a validated pipeline combining nonparametric association tests and stability audits to improve the reliability of AI-driven risk assessment in food chemistry.
Using artificial intelligence tools for data quality evaluation in the context of microplastic human health risk assessments
AI summary Read the abstract
Researchers tested whether AI tools like ChatGPT and Gemini could help evaluate the quality of microplastics research studies, a task that is normally slow and inconsistent when done manually. The AI tools proved effective at extracting information and assessing study reliability, which could speed up the process of identifying trustworthy data for assessing how microplastics affect human health.
DRAFT-PATCH: a multi-pass large-language-model protocol for extracting monitoring data from the primary literature
AI summary Read the abstract
Researchers built an AI tool called DRAFT-PATCH that automatically reads scientific papers and pulls out data on contaminants like "forever chemicals" (PFAS), pesticides, and microplastics, work that normally takes humans months or years to do by hand. By checking its own work multiple times, the tool becomes accurate enough to scan thousands of studies at once, which could help scientists and regulators spot pollution patterns and gaps in monitoring much faster than before. This matters because better, faster tracking of harmful chemicals in our environment is a key step toward understanding and reducing our exposure to them.
A provenance-aware semantic framework and evidence-aware knowledge graph for micro- and nanoplastic environmental-health evidence
AI summary Read the abstract
Researchers built a computer system to organize and connect scientific studies on micro- and nanoplastics, covering topics like environmental spread, lab toxicity tests, and human exposure studies. The system aims to help link findings from very different types of research together in a structured, searchable way.
Research digests by email
When a large batch of papers lands in the Atlas, we read through it and send a short write-up of what stood out.