Known Unknowns
7 papers with code • 0 benchmarks • 0 datasets
Language models have a tendency to generate text containing false statements that are often referred to as "Hallucinations." The primary purpose of this task is to test for this failure case by probing whether a model can correctly identify that the answer to a question is unknown. A common failure mode would be to prefer a prediction of false on unknown truth over a prediction that the answer is unknown.
Source: BIG-bench
Benchmarks
These leaderboards are used to track progress in Known Unknowns
Most implemented papers
PaLM: Scaling Language Modeling with Pathways
lucidrains/CoCa-pytorch
•
•
Google Research 2022
To further our understanding of the impact of scale on few-shot learning, we trained a 540-billion parameter, densely activated, Transformer language model, which we call Pathways Language Model PaLM.
Generative ODE Modeling with Known Unknowns
orilinial/GOKU
•
•
ICLR Workshop DeepDiffEq 2019
A motivating example is intensive care unit patients: the dynamics of vital physiological functions, such as the cardiovascular system with its associated variables (heart rate, cardiac contractility and output and vascular resistance) can be approximately described by a known system of ODEs.
Scaling Language Models: Methods, Analysis & Insights from Training Gopher
Language modelling provides a step towards intelligent communication systems by harnessing large repositories of written human knowledge to better predict and understand the world.
Training Compute-Optimal Large Language Models
We investigate the optimal model size and number of tokens for training a transformer language model under a given compute budget.
Known Unknowns: Uncertainty Quality in Bayesian Neural Networks
We compare the following candidate neural network models: Maximum Likelihood, Bayesian Dropout, OSBA, and --- for MNIST --- the standard variational approximation.
The division of labor in communication: Speakers help listeners account for asymmetries in visual perspective
In Experiment 1, we manipulated the presence or absence of occlusions in a director-matcher task and found that speakers spontaneously produced more informative descriptions to account for "known unknowns" in their partner's private view.
High-dimensional forecasting with known knowns and known unknowns
Forecasts play a central role in decision making under uncertainty.
