Search Results for author:
Found 19 papers, 10 papers with code
Adaptive Testing and Debugging of NLP Models
Current approaches to testing and debugging NLP models rely on highly variable human creativity and extensive labor, or only work for a very restrictive class of bugs.
Sparks of Artificial General Intelligence: Early experiments with GPT-4
We contend that (this early version of) GPT-4 is part of a new cohort of LLMs (along with ChatGPT and Google's PaLM for example) that exhibit more general intelligence than previous AI models.
Ranked #32 on
Arithmetic Reasoning
on GSM8K
ART: Automatic multi-step reasoning and tool-use for large language models
We introduce Automatic Reasoning and Tool-use (ART), a framework that uses frozen LLMs to automatically generate intermediate reasoning steps as a program.
Adaptive Testing of Computer Vision Models
Vision models often fail systematically on groups of data that share common semantic characteristics (e. g., rare objects or unusual scenes), but identifying these failure modes is a challenge.
Fixing Model Bugs with Natural Language Patches
Current approaches for fixing systematic problems in NLP models (e. g. regex patches, finetuning on more data) are either brittle, or labor-intensive and liable to shortcuts.
Axiomatic Explanations for Visual Search, Retrieval, and Similarity Learning
Visual search, recommendation, and contrastive similarity learning power technologies that impact billions of users worldwide.
Explaining by Removing: A Unified Framework for Model Explanation
We describe a new unified class of methods, removal-based explanations, that are based on the principle of simulating feature removal to quantify each feature's influence.
Feature Removal Is a Unifying Principle for Model Explanation Methods
Researchers have proposed a wide variety of model explanation approaches, but it remains unclear how most methods are related or when one method is preferable to another.
Shapley Flow: A Graph-based Approach to Interpreting Model Predictions
A causal graph, which encodes the relationships among input variables, can aid in assigning feature importance.
True to the Model or True to the Data?
Furthermore, we argue that the choice comes down to whether it is desirable to be true to the model or true to the data.
Understanding Global Feature Contributions With Additive Importance Measures
Understanding the inner workings of complex machine learning models is a long-standing problem and most recent research has focused on local interpretability.
Forecasting adverse surgical events using self-supervised transfer learning for physiological signals
Here, we present a transferable embedding method (i. e., a method to transform time series signals into input features for predictive machine learning models) named PHASE (PHysiologicAl Signal Embeddings) that enables us to more accurately forecast adverse surgical outcomes based on physiological signals.
Explaining Models by Propagating Shapley Values of Local Components
In healthcare, making the best possible predictions with complex models (e. g., neural networks, ensembles/stacks of different models) can impact patient welfare.
Improving performance of deep learning models with axiomatic attribution priors and expected gradients
Recent research has demonstrated that feature attribution methods for deep networks can themselves be incorporated into training; these attribution priors optimize for a model whose attributions have certain desirable properties -- most frequently, that particular features are important or unimportant.
Physiological Signal Embeddings (PHASE) via Interpretable Stacked Models
Here, we present the PHASE (PHysiologicAl Signal Embeddings) framework, which consists of three components: i) learning neural network embeddings of physiological signals, ii) predicting outcomes based on the learned embedding, and iii) interpreting the prediction results by estimating feature attributions in the "stacked" models (i. e., feature embedding model followed by prediction model).
Hybrid Gradient Boosting Trees and Neural Networks for Forecasting Operating Room Data
In this paper, we present feature learning via long short term memory (LSTM) networks and prediction via gradient boosting trees (XGB).
Checkpoint Ensembles: Ensemble Methods from a Single Training Process
We present the checkpoint ensembles method that can learn ensemble models on a single training process.
A Unified Approach to Interpreting Model Predictions
Understanding why a model makes a certain prediction can be as crucial as the prediction's accuracy in many applications.
An unexpected unity among methods for interpreting model predictions
Here, we present how a model-agnostic additive representation of the importance of input features unifies current methods.
