This program is designed to provide decision-makers, policymakers and professional working in policy with a foundational understanding of AI technology.
We use cookies to analyze the browsing and usage of our website and to personalize your experience. You can disable these technologies at any time, but this may limit certain functionalities of the site. Read our Privacy Policy for more information.
Setting cookies
You can enable and disable the types of cookies you wish to accept. However certain choices you make could affect the services offered on our sites (e.g. suggestions, personalised ads, etc.).
Essential cookies
These cookies are necessary for the operation of the site and cannot be deactivated. (Still active)
Analytics cookies
Do you accept the use of cookies to measure the audience of our sites?
Multimedia Player
Do you accept the use of cookies to display and allow you to watch the video content hosted by our partners (YouTube, etc.)?
Publications
Bugs in Large Language Models Generated Code: An Empirical Study
OBJECTIVE
In bioelectronic medicine, neuromodulation therapies induce neural signals to the brain or organs, modifying their function. Stimu… (see more)lation devices capable of triggering exogenous neural signals using electrical waveforms require a complex and multi-dimensional parameter space to control such waveforms. Determining the best combination of parameters (waveform optimization or dosing) for treating a particular patient's illness is therefore challenging. Comprehensive parameter searching for an optimal stimulation effect is often infeasible in a clinical setting due to the size of the parameter space. Restricting this space, however, may lead to suboptimal therapeutic results, reduced responder rates, and adverse effects. Approach. As an alternative to a full parameter search, we present a flexible machine learning, data acquisition, and processing framework for optimizing neural stimulation parameters, requiring as few steps as possible using Bayesian optimization. This optimization builds a model of the neural and physiological responses to stimulations, enabling it to optimize stimulation parameters and provide estimates of the accuracy of the response model. The vagus nerve innervates, among other thoracic and visceral organs, the heart, thus controlling heart rate, making it an ideal candidate for demonstrating the effectiveness of our approach. Main results. The efficacy of our optimization approach was first evaluated on simulated neural responses, then applied to vagus nerve stimulation intraoperatively in porcine subjects. Optimization converged quickly on parameters achieving target heart rates and optimizing neural B-fiber activations despite high intersubject variability. Significance. An optimized stimulation waveform was achieved in real time with far fewer stimulations than required by alternative optimization strategies, thus minimizing exposure to side effects. Uncertainty estimates helped avoiding stimulations outside a safe range. Our approach shows that a complex set of neural stimulation parameters can be optimized in real-time for a patient to achieve a personalized precision dosing. .
We present ScatterMoE, an implementation of Sparse Mixture-of-Experts (SMoE) on GPUs. ScatterMoE builds upon existing implementations, and o… (see more)vercoming some of the limitations to improve inference and training speed, and memory footprint. This implementation achieves this by avoiding padding and making excessive copies of the input. We introduce ParallelLinear, the main component we use to build our implementation and the various kernels used to speed up the operation. We benchmark our implementation against Megablocks, and show that it enables a higher throughput and lower memory footprint. We also show how ParallelLinear enables extension of the Mixture-of-Experts concept by demonstrating with an implementation of Mixture of Attention.
Large language models (LLMs) are routinely pre-trained on billions of tokens, only to start the process over again once new data becomes ava… (see more)ilable. A much more efficient solution is to continually pre-train these models, saving significant compute compared to re-training. However, the distribution shift induced by new data typically results in degraded performance on previous data or poor adaptation to the new data. In this work, we show that a simple and scalable combination of learning rate (LR) re-warming, LR re-decaying, and replay of previous data is sufficient to match the performance of fully re-training from scratch on all available data, as measured by the final loss and the average score on several language model (LM) evaluation benchmarks. Specifically, we show this for a weak but realistic distribution shift between two commonly used LLM pre-training datasets (English
We study the use of large language model-based agents for interacting with software via web browsers. Unlike prior work, we focus on measuri… (see more)ng the agents' ability to perform tasks that span the typical daily work of knowledge workers utilizing enterprise software systems. To this end, we propose WorkArena, a remote-hosted benchmark of 29 tasks based on the widely-used ServiceNow platform. We also introduce BrowserGym, an environment for the design and evaluation of such agents, offering a rich set of actions as well as multimodal observations. Our empirical evaluation reveals that while current agents show promise on WorkArena, there remains a considerable gap towards achieving full task automation. Notably, our analysis uncovers a significant performance disparity between open and closed-source LLMs, highlighting a critical area for future exploration and development in the field.
While Transformers have enabled tremendous progress in various application settings, such architectures still lag behind traditional symboli… (see more)c planners for solving complex decision making tasks.
In this work, we demonstrate how to train Transformers to solve complex planning tasks.
This is accomplished by training an encoder-decoder Transformer model to predict the _search dynamics_ of the
Continuous normalizing flows (CNFs) are an attractive generative modeling technique, but they have been held back by limitations in their si… (see more)mulation-based maximum likelihood training. We introduce the generalized \textit{conditional flow matching} (CFM) technique, a family of simulation-free training objectives for CNFs. CFM features a stable regression objective like that used to train the stochastic flow in diffusion models but enjoys the efficient inference of deterministic flow models. In contrast to both diffusion models and prior CNF training algorithms, CFM does not require the source distribution to be Gaussian or require evaluation of its density. A variant of our objective is optimal transport CFM (OT-CFM), which creates simpler flows that are more stable to train and lead to faster inference, as evaluated in our experiments. Furthermore, OT-CFM is the first method to compute dynamic OT in a simulation-free way. Training CNFs with CFM improves results on a variety of conditional and unconditional generation tasks, such as inferring single cell dynamics, unsupervised image translation, and Schrödinger bridge inference.