**Laconic**

We motivate a method for transparently identifying ineffectual computations in unmodified Deep Learning models and without affecting accuracy. Specifically, we show that if we decompose multiplications down to the bit level the amount of work performed during inference for image classification models can be consistently reduced by two orders of magnitude. In the best case studied of a sparse variant of AlexNet, this approach can ideally reduce computation work by more than 500x. We present Laconic a hardware accelerator that implements this approach to improve execution time, and energy efficiency for inference with Deep Learning Networks. Laconic judiciously gives up some of the work reduction potential to yield a low-cost, simple, and energy efficient design that outperforms other state-of-the-art accelerators. For example, a Laconic configuration that uses a weight memory interface with just 128 wires outperforms a conventional accelerator with a 2K-wire weight memory interface by 2.3x on average while being 2.13x more energy efficient on average. A Laconic configuration that uses a 1K-wire weight memory interface, outperforms the 2K-wire conventional accelerator by 15.4x and is 1.95x more energy efficient. Laconic does not require but rewards advances in model design such as a reduction in precision, the use of alternate numeric representations that reduce the number of bits that are ‘1’, or an increase in weight or activation sparsity. … **IPOC**

The performance of a reinforcement learning algorithm can vary drastically during learning because of exploration. Existing algorithms provide little information about their current policy’s quality before executing it, and thus have limited use in high-stakes applications like healthcare. In this paper, we address such a lack of accountability by proposing that algorithms output policy certificates, which upper bound the suboptimality in the next episode, allowing humans to intervene when the certified quality is not satisfactory. We further present a new learning framework (IPOC) for finite-sample analysis with policy certificates, and develop two IPOC algorithms that enjoy guarantees for the quality of both their policies and certificates. … **Stochastic Gradient Tree**

We present an online algorithm that induces decision trees using gradient information as the source of supervision. In contrast to previous approaches to gradient-based tree learning, we do not require soft splits or construction of a new tree for every update. In experiments, our method performs comparably to standard incremental classification trees and outperforms state of the art incremental regression trees. We also show how the method can be used to construct a novel type of neural network layer suited to learning representations from tabular data and find that it increases accuracy of multiclass and multi-label classification. … **Neutrosophic Logic**

A logic in which each proposition is estimated to have the percentage of truth in a subset T, the percentage of indeterminacy in a subset I, and the percentage of falsity in a subset F, is called Neutrosophic Logic. We use a subset of truth (or indeterminacy, or falsity), instead of a number only, because in many cases we are not able to exactly determine the percentages of truth and of falsity but to approximate them: for example a proposition is between 30-40% true and between 60-70% false, even worst: between 30-40% or 45-50% true (according to various analyzers), and 60% or between 66-70% false. The subsets are not necessary intervals, but any sets (discrete, continuous, open or closed or halfopen/ half-closed interval, intersections or unions of the previous sets, etc.) in accordance with the given proposition. A subset may have one element only in special cases of this logic. …

# If you did not already know

**21**
*Monday*
Mar 2022

Posted What is ...

in