Foundation Models
Foundation models serve as a backbone for diverse AI applications. How are they changing fields like natural language processing, vision, and robotics?
Small models get better, regulation moves to the states, and more.
Small models get better, regulation moves to the states, and more.
Current societal trends reflect an increased mistrust in science and a lowered civic engagement that threaten to impair research that is foundational for ensuring public health and advancing health equity. One effective countermeasure to these trends lies in community-facing citizen science applications to increase public participation in scientific research, making this field an important target for artificial intelligence (AI) exploration. We highlight potentially promising citizen science AI applications that extend beyond individual use to the community level, including conversational large language models, text-to-image generative AI tools, descriptive analytics for analyzing integrated macro- and micro-level data, and predictive analytics. The novel adaptations of AI technologies for community-engaged participatory research also bring an array of potential risks. We highlight possible negative externalities and mitigations for some of the potential ethical and societal challenges in this field.
Current societal trends reflect an increased mistrust in science and a lowered civic engagement that threaten to impair research that is foundational for ensuring public health and advancing health equity. One effective countermeasure to these trends lies in community-facing citizen science applications to increase public participation in scientific research, making this field an important target for artificial intelligence (AI) exploration. We highlight potentially promising citizen science AI applications that extend beyond individual use to the community level, including conversational large language models, text-to-image generative AI tools, descriptive analytics for analyzing integrated macro- and micro-level data, and predictive analytics. The novel adaptations of AI technologies for community-engaged participatory research also bring an array of potential risks. We highlight possible negative externalities and mitigations for some of the potential ethical and societal challenges in this field.
This brief presents an analysis of Chinese AI startup DeepSeek’s talent base and calls for U.S. policymakers to reinvest in competing to attract and retain global AI talent.
This brief presents an analysis of Chinese AI startup DeepSeek’s talent base and calls for U.S. policymakers to reinvest in competing to attract and retain global AI talent.
.png&w=256&q=80)
.png&w=1920&q=100)
The new AI Index is out. See how well you know the state of the industry.
The new AI Index is out. See how well you know the state of the industry.
Model-based reinforcement learning (MBRL) is a promising route to sampleefficient policy optimization. However, a known vulnerability of reconstructionbased MBRL consists of scenarios in which detailed aspects of the world are highly predictable, but irrelevant to learning a good policy. Such scenarios can lead the model to exhaust its capacity on meaningless content, at the cost of neglecting important environment dynamics. While existing approaches attempt to solve this problem, we highlight its continuing impact on leading MBRL methods —including DreamerV3 and DreamerPro — with a novel environment where background distractions are intricate, predictable, and useless for planning future actions. To address this challenge we develop a method for focusing the capacity of the world model through synergy of a pretrained segmentation model, a task-aware reconstruction loss, and adversarial learning. Our method outperforms a variety of other approaches designed to reduce the impact of distractors, and is an advance towards robust model-based reinforcement learning.
Model-based reinforcement learning (MBRL) is a promising route to sampleefficient policy optimization. However, a known vulnerability of reconstructionbased MBRL consists of scenarios in which detailed aspects of the world are highly predictable, but irrelevant to learning a good policy. Such scenarios can lead the model to exhaust its capacity on meaningless content, at the cost of neglecting important environment dynamics. While existing approaches attempt to solve this problem, we highlight its continuing impact on leading MBRL methods —including DreamerV3 and DreamerPro — with a novel environment where background distractions are intricate, predictable, and useless for planning future actions. To address this challenge we develop a method for focusing the capacity of the world model through synergy of a pretrained segmentation model, a task-aware reconstruction loss, and adversarial learning. Our method outperforms a variety of other approaches designed to reduce the impact of distractors, and is an advance towards robust model-based reinforcement learning.
All Work Published on Foundation Models
Stanford HAI researchers create eight new AI benchmarks that could help developers reduce bias in AI models, potentially making them fairer and less likely to case harm.
Stanford HAI researchers create eight new AI benchmarks that could help developers reduce bias in AI models, potentially making them fairer and less likely to case harm.
Vafa et al. (2024) introduced a transformer-based econometric model, CAREER, that predicts a worker’s next job as a function of career history (an “occupation model”). CAREER was initially estimated (“pre-trained”) using a large, unrepresentative resume dataset, which served as a “foundation model,” and parameter estimation was continued (“fine-tuned”) using data from a representative survey. CAREER had better predictive performance than benchmarks. This paper considers an alternative where the resume-based foundation model is replaced by a large language model (LLM). We convert tabular data from the survey into text files that resemble resumes and fine-tune the LLMs using these text files with the objective to predict the next token (word). The resulting fine-tuned LLM is used as an input to an occupation model. Its predictive performance surpasses all prior models. We demonstrate the value of fine-tuning and further show that by adding more career data from a different population, fine-tuning smaller LLMs surpasses the performance of fine-tuning larger models.
Vafa et al. (2024) introduced a transformer-based econometric model, CAREER, that predicts a worker’s next job as a function of career history (an “occupation model”). CAREER was initially estimated (“pre-trained”) using a large, unrepresentative resume dataset, which served as a “foundation model,” and parameter estimation was continued (“fine-tuned”) using data from a representative survey. CAREER had better predictive performance than benchmarks. This paper considers an alternative where the resume-based foundation model is replaced by a large language model (LLM). We convert tabular data from the survey into text files that resemble resumes and fine-tune the LLMs using these text files with the objective to predict the next token (word). The resulting fine-tuned LLM is used as an input to an occupation model. Its predictive performance surpasses all prior models. We demonstrate the value of fine-tuning and further show that by adding more career data from a different population, fine-tuning smaller LLMs surpasses the performance of fine-tuning larger models.
What Makes a Good AI Benchmark?
This brief presents a novel assessment framework for evaluating the quality of AI benchmarks and scores 24 benchmarks against the framework.
This brief presents a novel assessment framework for evaluating the quality of AI benchmarks and scores 24 benchmarks against the framework.
A study led by Stanford HAI Faculty Fellow Johannes Eichstaedt reveals that large language models adapt their behavior to appear more likable when they are being studied, mirroring human tendencies to present favorably.
A study led by Stanford HAI Faculty Fellow Johannes Eichstaedt reveals that large language models adapt their behavior to appear more likable when they are being studied, mirroring human tendencies to present favorably.
A Multi-Center Study on the Adaptability of a Shared Foundation Model for Electronic Health Records
Foundation models are transforming artificial intelligence (AI) in healthcare by providing modular components adaptable for various downstream tasks, making AI development more scalable and cost-effective. Foundation models for structured electronic health records (EHR), trained on coded medical records from millions of patients, demonstrated benefits including increased performance with fewer training labels, and improved robustness to distribution shifts. However, questions remain on the feasibility of sharing these models across hospitals and their performance in local tasks. This multi-center study examined the adaptability of a publicly accessible structured EHR foundation model (FMSM), trained on 2.57 M patient records from Stanford Medicine. Experiments used EHR data from The Hospital for Sick Children (SickKids) and Medical Information Mart for Intensive Care (MIMIC-IV). We assessed both adaptability via continued pretraining on local data, and task adaptability compared to baselines of locally training models from scratch, including a local foundation model. Evaluations on 8 clinical prediction tasks showed that adapting the off-the-shelf FMSMmatched the performance of gradient boosting machines (GBM) locally trained on all data while providing a 13% improvement in settings with few task-specific training labels. Continued pretraining on local data showed FMSM required fewer than 1% of training examples to match the fully trained GBM’s performance, and was 60 to 90% more sample-efficient than training local foundation models from scratch. Our findings demonstrate that adapting EHR foundation models across hospitals provides improved prediction performance at less cost, underscoring the utility of base foundation models as modular components to streamline the development of healthcare AI.
Foundation models are transforming artificial intelligence (AI) in healthcare by providing modular components adaptable for various downstream tasks, making AI development more scalable and cost-effective. Foundation models for structured electronic health records (EHR), trained on coded medical records from millions of patients, demonstrated benefits including increased performance with fewer training labels, and improved robustness to distribution shifts. However, questions remain on the feasibility of sharing these models across hospitals and their performance in local tasks. This multi-center study examined the adaptability of a publicly accessible structured EHR foundation model (FMSM), trained on 2.57 M patient records from Stanford Medicine. Experiments used EHR data from The Hospital for Sick Children (SickKids) and Medical Information Mart for Intensive Care (MIMIC-IV). We assessed both adaptability via continued pretraining on local data, and task adaptability compared to baselines of locally training models from scratch, including a local foundation model. Evaluations on 8 clinical prediction tasks showed that adapting the off-the-shelf FMSMmatched the performance of gradient boosting machines (GBM) locally trained on all data while providing a 13% improvement in settings with few task-specific training labels. Continued pretraining on local data showed FMSM required fewer than 1% of training examples to match the fully trained GBM’s performance, and was 60 to 90% more sample-efficient than training local foundation models from scratch. Our findings demonstrate that adapting EHR foundation models across hospitals provides improved prediction performance at less cost, underscoring the utility of base foundation models as modular components to streamline the development of healthcare AI.
In this response to the U.S. AI Safety Institute’s (US AISI) request for comment on its draft guidelines for managing the misuse risk for dual-use foundation models, scholars from Stanford HAI, the Center for Research on Foundation Models (CRFM), and the Regulation, Evaluation, and Governance Lab (RegLab) urge the US AISI to strengthen its guidance on reproducible evaluations and third- party evaluations, as well as clarify guidance on post-deployment monitoring. They also encourage the institute to develop similar guidance for other actors in the foundation model supply chain and for non-misuse risks, while ensuring the continued open release of foundation models absent evidence of marginal risk.
In this response to the U.S. AI Safety Institute’s (US AISI) request for comment on its draft guidelines for managing the misuse risk for dual-use foundation models, scholars from Stanford HAI, the Center for Research on Foundation Models (CRFM), and the Regulation, Evaluation, and Governance Lab (RegLab) urge the US AISI to strengthen its guidance on reproducible evaluations and third- party evaluations, as well as clarify guidance on post-deployment monitoring. They also encourage the institute to develop similar guidance for other actors in the foundation model supply chain and for non-misuse risks, while ensuring the continued open release of foundation models absent evidence of marginal risk.
