Choosing fairly when you don't know how many you're choosing from.
#Exploration: A Study of Count-Based Exploration for Deep...
Count-based exploration algorithms are known to perform near-optimally when used in conjunction with tabular reinforcement learning (RL) methods for solving small discrete Markov decision...
Beware of samples! A cognitive-ecological sampling approach to judgment biases.
Opinion | This Is What Will Ruin Public Opinion Polling for Good
Instead of navigating the obstacles to conduct polls with human respondents, pollsters are running A.I. simulations instead. Why?
Adding Realistic Rivers to Random Terrain
Discusses a method for adding rivers to randomly generated terrain that adds a realistic touch.
Complete guide to samplers in Stable Diffusion
Dive into the world of Stable Diffusion samplers and unlock the potential of image generation.
AI Aqueducts: Speed. Confidence. And Contaminants.
The AI Pipeline Has No Way to Judge the Science It Uses
LAVA: Data Valuation without Pre-Specified Learning Algorithms
Traditionally, data valuation is posed as a problem of equitably splitting the validation performance of a learning algorithm among the training data. As a result, the calculated data values depend on many design choices of the underlying learning algorithm. However, this dependence is undesirable for many use cases of data valuation, such as setting priorities over different data sources in a data acquisition process and informing pricing mechanisms in a data marketplace. In these scenarios, data needs to be valued before the actual analysis and the choice of the learning algorithm is still undetermined then. Another side-effect of the dependence is that to assess the value of individual points, one needs to re-run the learning algorithm with and without a point, which incurs a large computation burden. This work leapfrogs over the current limits of data valuation methods by introducing a new framework that can value training data in a way that is oblivious to the downstream learning algorithm. Our main results are as follows. $\textbf{(1)}$ We develop a proxy for the validation performance associated with a training set based on a non-conventional $\textit{class-wise}$ $\textit{Wasserstein distance}$ between the training and the validation set. We show that the distance characterizes the upper bound of the validation performance for any given model under certain Lipschitz conditions. $\textbf{(2)}$ We develop a novel method to value individual data based on the sensitivity analysis of the $\textit{class-wise}$ Wasserstein distance. Importantly, these values can be directly obtained $\textit{for free}$ from the output of off-the-shelf optimization solvers once the Wasserstein distance is computed. $\textbf{(3) }$We evaluate our new data valuation framework over various use cases related to detecting low-quality data and show that, surprisingly, the learning-agnostic feature of our framework enables a $\textit{significant improvement}$ over the state-of-the-art performance while being $\textit{orders of magnitude faster.}$
Training a vision model, or how I learned to stop worrying and love dataset curation
Note: this is not some beginners guide, or how-to. Its an experiment in high dimensional space manifold gymnastics. Quick links to the model and dataset if you don’t care for the narrative version: This is the Badger Model 55 Water Meter: This thing sucks to read. Get down on your knees, enjoy the cement floor of a dark basement room, pull out a flash light and just try to guess if water is leaking from somewhere. The water company knows what they bill you, but your costs jump around li...
People Reject Algorithms in Uncertain Decision Domains Because They Have Diminishing Sensitivity to Forecasting Error
Will people use self-driving cars, virtual doctors, and other algorithmic decision-makers if they outperform humans? The answer depends on the uncertainty inherent in the decision domain. We propose that people have diminishing sensitivity to forecasting error and that this preference results in people favoring riskier (and often worse-performing) decision-making methods, such as human judgment, in inherently uncertain domains. In nine studies ( N = 4,820), we found that (a) people have diminishing sensitivity to each marginal unit of error that a forecast produces, (b) people are less likely to use the best possible algorithm in decision domains that are more unpredictable, (c) people choose between decision-making methods on the basis of the perceived likelihood of those methods producing a near-perfect answer, and (d) people prefer methods that exhibit higher variance in performance (all else being equal). To the extent that investing, medical decision-making, and other domains are inherently uncertain, people may be unwilling to use even the best possible algorithm in those domains.
Models
Amp uses the best model for each task: leading generalist foundation models for complex reasoning and planning, and smaller specialized models for fast, accurate responses in specific domains.
AI Sycophancy and Decisions
We examine whether sycophantic AI advice distorts decisions. Our experiment involves 1,500 participants in 30 decision environments spanning core domains in eco
