Search Results for author:
Found 6 papers, 4 papers with code
Out-of-Distribution Detection & Applications With Ablated Learned Temperature Energy
As deep neural networks become adopted in high-stakes domains, it is crucial to identify when inference inputs are Out-of-Distribution (OOD) so that users can be alerted of likely drops in performance and calibration despite high confidence -- ultimately to know when networks' decisions (and their uncertainty in those decisions) should be trusted.
A Baseline Analysis of Reward Models' Ability To Accurately Analyze Foundation Models Under Distribution Shift
These reward models are additionally used at inference-time to estimate LLM responses' adherence to those desired behaviors.
Enabling Calibration In The Zero-Shot Inference of Large Vision-Language Models
Calibration of deep learning models is crucial to their trustworthiness and safe usage, and as such, has been extensively studied in supervised classification models, with methods crafted to decrease miscalibration.
Deep Discriminative to Kernel Density Graph for In- and Out-of-distribution Calibrated Inference
However, these methods are not calibrated for the entire feature space, leading to overconfidence in the case of out-of-distribution (OOD) samples.
When are Deep Networks really better than Decision Forests at small sample sizes, and how?
Empirically, we compare these two strategies on hundreds of tabular data settings, as well as several vision and auditory settings.
Omnidirectional Transfer for Quasilinear Lifelong Learning
But striving to avoid forgetting sets the goal unnecessarily low: the goal of lifelong learning, whether biological or artificial, should be to improve performance on all tasks (including past and future) with any new data.
