A ManyModels project would be a valuable contribution. Almost all published literature using LLMs is out of date by time of peer-review; maybe a living report to show 'current' state of LLMs? What's the experimental design?

Fantastic, thanks for doing this and making it available!

Thanks Nick! Much appreciated :)

‼️"o1-preview demonstrates superhuman performance in differential diagnosis, diagnostic clinical reasoning, and management reasoning, superior in multiple domains compared to prior model generations and human physicians."

And this is using vignettes, not multiple choice. arxiv.org/pdf/2412.10849

Add me please :) 👋

this may end up becoming how evidence synthesis is done in the future (with teams of AI agents collaborating on each step of the process and continuously updating reviews/meta-analyses when new evidence becomes available)

Challenging indeed! Regression based models will only get us so far. I think formal models (eg ordinary differential equations) are needed at the idiographic level. I'm hopeful that personal AI agents will be developed to do this type of continual data collection and modeling in the future.

Clustering of Health Behaviors in Canadians: A Multiple Behavior Analysis of Data from the Canadian Longitudinal Study on Aging

academic.oup.com/abm/article/...

Effectiveness of Interventions for Changing More Than One Behavior at a Time to Manage Chronic Conditions: A Systematic Review and Meta-analysis:
academic.oup.com/abm/article/...

Components of multiple health behaviour change interventions for patients with chronic conditions: a systematic review and meta-regression of randomized trials:
www.tandfonline.com/doi/full/10....

barriers/enablers of attendance at eye screening among three groups of immigrants to Canada from cultural/linguistic minority groups living with diabetes.

onlinelibrary.wiley.com/doi/10.1111/...

Functional dependence can be best prospectively estimated by age, psychological distress, physical fitness, physical activity, chronic conditions, and sex. These predictors can
estimate functional dependence more than 6 years in advance with high accuracy.

www.medrxiv.org/content/10.1...

The beneficial effect of prestroke physical activity on activities of daily living (ADL) limitations after stroke is stronger than its effect in matched adults without stroke followed for a similar number of years.

@matthieuboisgontier.com

academic.oup.com/ptj/article/...

Using network analysis to examine the temporal dynamics of multiple health behaviours and pandemic health behaviours

@jpresseau.bsky.social

bpspsychub.onlinelibrary.wiley.com/doi/full/10....