Search Results for author:
Found 19 papers, 9 papers with code
Societal Alignment Frameworks Can Improve LLM Alignment
Recent progress in large language models (LLMs) has focused on producing responses that meet human expectations and align with shared values - a process coined alignment.
Monte Carlo Tree Search Boosts Reasoning via Iterative Preference Learning
We introduce an approach aimed at enhancing the reasoning capabilities of Large Language Models (LLMs) through an iterative preference learning process inspired by the successful strategy employed by AlphaZero.
Towards Biologically Plausible Convolutional Networks
This requires the network to pause occasionally for a sleep-like phase of "weight sharing".
Training Generative Adversarial Networks by Solving Ordinary Differential Equations
From this perspective, we hypothesise that instabilities in training GANs arise from the integration error in discretising the continuous dynamics.
Compressive Transformers for Long-Range Sequence Modelling
We present the Compressive Transformer, an attentive sequence model which compresses past memories for long-range sequence learning.
Ranked #2 on
Language Modelling
on Hutter Prize
Meta-Learning Deep Energy-Based Memory Models
We study the problem of learning associative memory -- a system which is able to retrieve a remembered pattern based on its distorted or incomplete version.
Automated curricula through setter-solver interactions
We demonstrate the success of our approach in rich but sparsely rewarding 2D and 3D environments, where an agent is tasked to achieve a single goal selected from a set of possible goals that varies between episodes, and identify challenges for future work.
What does it mean to understand a neural network?
In analogy, we conjecture that rules for development and learning in brains may be far easier to understand than their resulting properties.
Meta-Learning Neural Bloom Filters
There has been a recent trend in training neural networks to replace data structures that have been crafted by hand, with an aim for faster execution, better accuracy, or greater compression.
Composing Entropic Policies using Divergence Correction
Composing previously mastered skills to solve novel tasks promises dramatic improvements in the data efficiency of reinforcement learning.
Experience Replay for Continual Learning
We examine this issue in the context of reinforcement learning, in a setting where an agent is exposed to tasks in a sequence.
Fast Parametric Learning with Activation Memorization
Neural networks trained with backpropagation often struggle to identify classes that have been observed a small number of times.
Ranked #68 on
Language Modelling
on WikiText-103
Learning to Learn without Gradient Descent by Gradient Descent
We learn recurrent neural network optimizers trained on simple synthetic functions by gradient descent.
Scaling Memory-Augmented Neural Networks with Sparse Reads and Writes
SAM learns with comparable data efficiency to existing models on a range of synthetic tasks and one-shot Omniglot character recognition, and can scale to tasks requiring $100,\! 000$s of time steps and memories.
Ranked #6 on
Question Answering
on bAbi
(Mean Error Rate metric)
Towards deep learning with segregated dendrites
Deep learning has led to significant advances in artificial intelligence, in part, by adopting strategies motivated by neurophysiology.
Asynchronous Methods for Deep Reinforcement Learning
We propose a conceptually simple and lightweight framework for deep reinforcement learning that uses asynchronous gradient descent for optimization of deep neural network controllers.
Ranked #9 on
Atari Games
on Atari 2600 Star Gunner
Memory-based control with recurrent neural networks
Partially observed control problems are a challenging aspect of reinforcement learning.
Continuous control with deep reinforcement learning
We adapt the ideas underlying the success of Deep Q-Learning to the continuous action domain.
Ranked #2 on
OpenAI Gym
on HalfCheetah-v4
Random feedback weights support learning in deep neural networks
In machine learning, the backpropagation algorithm assigns blame to a neuron by computing exactly how it contributed to an error.
