Faster sorting algorithms discovered using deep reinforcement learning
Fundamental algorithms such as sorting or hashing are used trillions of times on any given
day. As demand for computation grows, it has become critical for these algorithms to be as …
day. As demand for computation grows, it has become critical for these algorithms to be as …
Combinatorial optimization and reasoning with graph neural networks
Combinatorial optimization is a well-established area in operations research and computer
science. Until recently, its methods have focused on solving problem instances in isolation …
science. Until recently, its methods have focused on solving problem instances in isolation …
EvoPrompting: language models for code-level neural architecture search
Given the recent impressive accomplishments of language models (LMs) for code
generation, we explore the use of LMs as general adaptive mutation and crossover …
generation, we explore the use of LMs as general adaptive mutation and crossover …
Randomized positional encodings boost length generalization of transformers
Transformers have impressive generalization capabilities on tasks with a fixed context
length. However, they fail to generalize to sequences of arbitrary length, even for seemingly …
length. However, they fail to generalize to sequences of arbitrary length, even for seemingly …
Parsel🐍: Algorithmic Reasoning with Language Models by Composing Decompositions
E Zelikman, Q Huang, G Poesia… - Advances in …, 2023 - proceedings.neurips.cc
Despite recent success in large language model (LLM) reasoning, LLMs struggle with
hierarchical multi-step reasoning tasks like generating complex programs. For these tasks …
hierarchical multi-step reasoning tasks like generating complex programs. For these tasks …
Neural algorithmic reasoning with causal regularisation
B Bevilacqua, K Nikiforou, B Ibarz… - International …, 2023 - proceedings.mlr.press
Recent work on neural algorithmic reasoning has investigated the reasoning capabilities of
neural networks, effectively demonstrating they can learn to execute classical algorithms on …
neural networks, effectively demonstrating they can learn to execute classical algorithms on …
LINC: A neurosymbolic approach for logical reasoning by combining language models with first-order logic provers
Logical reasoning, ie, deductively inferring the truth value of a conclusion from a set of
premises, is an important task for artificial intelligence with wide potential impacts on …
premises, is an important task for artificial intelligence with wide potential impacts on …
Multi-task learning for routing problem with cross-problem zero-shot generalization
Vehicle routing problems (VRP) are very important in many real-world applications and has
been studied for several decades. Recently, neural combinatorial optimization (NCO) has …
been studied for several decades. Recently, neural combinatorial optimization (NCO) has …
Towards better out-of-distribution generalization of neural algorithmic reasoning tasks
In this paper, we study the OOD generalization of neural algorithmic reasoning tasks, where
the goal is to learn an algorithm (eg, sorting, breadth-first search, and depth-first search) …
the goal is to learn an algorithm (eg, sorting, breadth-first search, and depth-first search) …
Asynchronous algorithmic alignment with cocycles
State-of-the-art neural algorithmic reasoners make use of message passing in graph neural
networks (GNNs). But typical GNNs blur the distinction between the definition and invocation …
networks (GNNs). But typical GNNs blur the distinction between the definition and invocation …