Search Results for author:
Found 83 papers, 30 papers with code
Accelerating Greedy Coordinate Gradient via Probe Sampling
Safety of Large Language Models (LLMs) has become a central issue given their rapid progress and wide applications.
Keeping LLMs Aligned After Fine-tuning: The Crucial Role of Prompt Templates
Public LLMs such as the Llama 2-Chat have driven huge activity in LLM research.
Can AI Be as Creative as Humans?
With the rise of advanced generative AI models capable of tasks once reserved for human creativity, the study of AI's creative potential becomes imperative for its responsible development and application.
Gemini: A Family of Highly Capable Multimodal Models
This report introduces a new family of multimodal models, Gemini, that exhibit remarkable capabilities across image, audio, video, and text understanding.
Ranked #1 on
Multi-task Language Understanding
on MMLU
(using extra training data)
Unlearning via Sparse Representations
Machine \emph{unlearning}, which involves erasing knowledge about a \emph{forget set} from a trained model, can prove to be costly and infeasible by existing techniques.
Physical Reasoning and Object Planning for Household Embodied Agents
In this study, we explore the sophisticated domain of task planning for robust household embodied agents, with a particular emphasis on the intricate task of selecting substitute objects.
Skill-Mix: a Flexible and Expandable Family of Evaluations for AI models
The paper develops a methodology for (a) designing and administering such an evaluation, and (b) automatic grading (plus spot-checking by humans) of the results using GPT-4 as well as the open LLaMA-2 70B model.
Aligning Text-to-Image Diffusion Models with Reward Backpropagation
Due to their unsupervised training, controlling their behavior in downstream tasks, such as maximizing human-perceived image quality, image-text alignment, or ethical image generation, is difficult.
A Theory for Emergence of Complex Skills in Language Models
Contributions include: (a) A statistical framework that relates cross-entropy loss of LLMs to competence on the basic skills that underlie language tasks.
Cycle Consistency Driven Object Discovery
To tackle the second limitation, we apply the learned object-centric representations from the proposed method to two downstream reinforcement learning tasks, demonstrating considerable performance enhancements compared to conventional slot-based and monolithic representation learning methods.
Spotlight Attention: Robust Object-Centric Learning With a Spatial Locality Prior
The aim of object-centric vision is to construct an explicit representation of the objects in a scene.
DiscoGen: Learning to Discover Gene Regulatory Networks
To accurately identify GRNs, perturbational data is required.
TC-VAE: Uncovering Out-of-Distribution Data Generative Factors
We show that the proposed model is able to uncover OOD generative factors on different datasets and outperforms on average the related baselines in terms of downstream disentanglement metrics.
Leveraging the Third Dimension in Contrastive Learning
We evaluate these two approaches on three different SSL methods -- BYOL, SimSiam, and SwAV -- using ImageNette (10 class subset of ImageNet), ImageNet-100 and ImageNet-1k datasets.
Representation Learning in Deep RL via Discrete Information Bottleneck
Several self-supervised representation learning methods have been proposed for reinforcement learning (RL) with rich observations.
Discrete Factorial Representations as an Abstraction for Goal Conditioned Reinforcement Learning
Goal-conditioned reinforcement learning (RL) is a promising direction for training agents that are capable of solving multiple tasks and reach a diverse set of objectives.
GFlowOut: Dropout with Generative Flow Networks
These methods face two important challenges: (a) the posterior distribution over masks can be highly multi-modal which can be difficult to approximate with standard variational inference and (b) it is not trivial to fully utilize sample-dependent information and correlation among dropout masks to improve posterior estimation.
Stateful active facilitator: Coordination and Environmental Heterogeneity in Cooperative Multi-Agent Reinforcement Learning
We formalize the notions of coordination level and heterogeneity level of an environment and present HECOGrid, a suite of multi-agent RL environments that facilitates empirical evaluation of different MARL approaches across different levels of coordination and environmental heterogeneity by providing a quantitative control over coordination and heterogeneity levels of the environment.
Multi-agent Reinforcement Learning
reinforcement-learning
+1
Discrete Key-Value Bottleneck
Deep neural networks perform well on classification tasks where data streams are i. i. d.
On the Generalization and Adaption Performance of Causal Models
Our analysis shows that the modular neural causal models outperform other models on both zero and few-shot adaptation in low data regimes and offer robust generalization.
Temporal Latent Bottleneck: Synthesis of Fast and Slow Processing Mechanisms in Sequence Learning
A slow stream that is recurrent in nature aims to learn a specialized and compressed representation, by forcing chunks of $K$ time steps into a single representation which is divided into multiple vectors.
Coordinating Policies Among Multiple Agents via an Intelligent Communication Channel
In Multi-Agent Reinforcement Learning (MARL), specialized channels are often introduced that allow agents to communicate directly with one another.
Intelligent Communication
Multi-agent Reinforcement Learning
+2
Learning to Induce Causal Structure
The fundamental challenge in causal induction is to infer the underlying graph structure given observational and/or interventional data.
Test-time Adaptation with Slot-Centric Models
In our work, we find evidence that these losses are insufficient for the task of scene decomposition, without also considering architectural inductive biases.
Retrieval-Augmented Reinforcement Learning
In this paper we explore an alternative paradigm in which we train a network to map a dataset of past experiences to optimal behavior.
Learning by Directional Gradient Descent
How should state be constructed from a sequence of observations, so as to best achieve some objective?
Learning Neural Causal Models with Active Interventions
Discovering causal structures from data is a challenging inference problem of fundamental importance in all areas of science.
Discrete-Valued Neural Communication
Deep learning has advanced from fully connected architectures to structured models organized into components, e. g., the transformer composed of positional elements, modular architectures divided into slots, and graph neural nets made up of nodes.
Systematic Evaluation of Causal Discovery in Visual Model Based Reinforcement Learning
A central goal for AI and causality is thus the joint discovery of abstract representations and causal structure.
Variational Causal Networks: Approximate Bayesian Inference over Causal Structures
However, a crucial aspect to acting intelligently upon the knowledge about causal structure which has been inferred from finite data demands reasoning about its uncertainty.
Robust Representation Learning via Perceptual Similarity Metrics
A fundamental challenge in artificial intelligence is learning useful representations of data that yield good performance on a downstream task, without overfitting to spurious input features.
Fast and Slow Learning of Recurrent Independent Mechanisms
To study these ideas, we propose a particular training framework in which we assume that the pieces of knowledge an agent needs and its reward function are stationary and can be re-used across tasks.
Neural Production Systems: Learning Rule-Governed Visual Dynamics
First, GNNs do not predispose interactions to be sparse, as relationships among independent entities are likely to be.
Coordination Among Neural Modules Through a Shared Global Workspace
We explore the use of such a communication channel in the context of deep learning for modeling the structure of complex environments.
Transformers with Competitive Ensembles of Independent Mechanisms
In this work we explore a way in which the Transformer architecture is deficient: it represents each position with a large monolithic hidden representation and a single set of parameters which are applied over the entire hidden representation.
Towards Causal Representation Learning
The two fields of machine learning and graphical causality arose and developed separately.
Meta Attention Networks: Meta-Learning Attention to Modulate Information Between Recurrent Independent Mechanisms
Decomposing knowledge into interchangeable pieces promises a generalization advantage when there are changes in distribution.
Factorizing Declarative and Procedural Knowledge in Structured, Dynamical Environments
To use a video game as an illustration, two enemies of the same type will share schemata but will have separate object files to encode their distinct state (e. g., health, position).
Learning Task-Relevant Features via Contrastive Input Morphing
A fundamental challenge in artificial intelligence is learning useful representations of data that yield good performance on a downstream classification task, without overfitting to spurious input features.
Dependency Structure Discovery from Interventions
Promising results have driven a recent surge of interest in continuous optimization methods for Bayesian network structure learning from observational data.
Spatially Structured Recurrent Modules
Capturing the structure of a data-generating process by means of appropriate inductive biases can help in learning models that generalise well and are robust to changes in the input distribution.
Inductive Biases for Deep Learning of Higher-Level Cognition
A fascinating hypothesis is that human and animal intelligence could be explained by a few principles (rather than an encyclopedic list of heuristics).
Neural Function Modules with Sparse Arguments: A Dynamic Approach to Integrating Information across Layers
Feed-forward neural networks consist of a sequence of layers, in which each layer performs some processing on the information from the previous layer.
CausalWorld: A Robotic Manipulation Benchmark for Causal Structure and Transfer Learning
To facilitate research addressing this problem, we propose CausalWorld, a benchmark for causal structure and transfer learning in a robotic manipulation environment.
S2RMs: Spatially Structured Recurrent Modules
Capturing the structure of a data-generating process by means of appropriate inductive biases can help in learning models that generalize well and are robust to changes in the input distribution.
Uniform Priors for Data-Efficient Transfer
Deep Neural Networks have shown great promise on a variety of downstream applications; but their ability to adapt and generalize to new data and tasks remains a challenge.
Learning to Combine Top-Down and Bottom-Up Signals in Recurrent Neural Networks with Attention over Modules
To effectively utilize the wealth of potential top-down information available, and to prevent the cacophony of intermixed signals in a bidirectional architecture, mechanisms are needed to restrict information flow.
Object Files and Schemata: Factorizing Declarative and Procedural Knowledge in Dynamical Systems
To use a video game as an illustration, two enemies of the same type will share schemata but will have separate object files to encode their distinct state (e. g., health, position).
Untangling tradeoffs between recurrence and self-attention in neural networks
Attention and self-attention mechanisms, are now central to state-of-the-art deep learning on sequential tasks.
On Disentangled Representations Learned From Correlated Data
The focus of disentanglement approaches has been on identifying independent factors of variation in data.
Learning the Arrow of Time for Problems in Reinforcement Learning
We humans have an innate understanding of the asymmetric progression of time, which we use to efficiently and safely perceive and manipulate our environment.
The Variational Bandwidth Bottleneck: Stochastic Evaluation on an Information Budget
This is typically the case when we have a standard conditioning input, such as a state observation, and a "privileged" input, which might correspond to the goal of a task, the output of a costly planning algorithm, or communication with another agent.
Diversity inducing Information Bottleneck in Model Ensembles
Although deep learning models have achieved state-of-the-art performance on a number of vision tasks, generalization over high dimensional multi-modal data, and reliable predictive uncertainty estimation are still active areas of research.
Top-k Training of GANs: Improving GAN Performance by Throwing Away Bad Samples
We introduce a simple (one line of code) modification to the Generative Adversarial Network (GAN) training algorithm that materially improves results with no increase in computational cost: When updating the generator parameters, we simply zero out the gradient contributions from the elements of the batch that the critic scores as `least realistic'.
Small-GAN: Speeding Up GAN Training Using Core-sets
Recent work by Brock et al. (2018) suggests that Generative Adversarial Networks (GANs) benefit disproportionately from large mini-batch sizes.
Learning Neural Causal Models from Unknown Interventions
Promising results have driven a recent surge of interest in continuous optimization methods for Bayesian network structure learning from observational data.
Recurrent Independent Mechanisms
Learning modular structures which reflect the dynamics of the environment can lead to better generalization and robustness to changes which only affect a few of the underlying causes.
Learning the Arrow of Time
We humans seem to have an innate understanding of the asymmetric progression of time, which we use to efficiently and safely perceive and manipulate our environment.
Reinforcement Learning with Competitive Ensembles of Information-Constrained Primitives
Reinforcement learning agents that operate in diverse and complex environments can benefit from the structured decomposition of their behavior.
Hierarchical Reinforcement Learning
reinforcement-learning
+1
Learning Powerful Policies by Using Consistent Dynamics Model
There is enough evidence that humans build a model of the environment, not only by observing the environment but also by interacting with the environment.
State-Reification Networks: Improving Generalization by Modeling the Distribution of Hidden Representations
Machine learning promises methods that generalize well from finite labeled data.
EnGAN: Latent Space MCMC and Maximum Entropy Generators for Energy-based Models
Unsupervised learning is about capturing dependencies between variables and is driven by the contrast between the probable vs improbable configurations of these variables, often either via a generative model which only samples probable ones or with an energy function (unnormalized log-density) which is low for probable ones and high for improbable ones.
Modeling the Long Term Future in Model-Based Reinforcement Learning
This paper focuses on building a model that reasons about the long-term future and demonstrates how to use this for efficient planning and exploration.
Transfer and Exploration via the Information Bottleneck
In new environments, this model can then identify novel subgoals for further exploration, guiding the agent through a sequence of potential decision states and through new regions of the state space.
Learning Dynamics Model in Reinforcement Learning by Incorporating the Long Term Future
This paper focuses on building a model that reasons about the long-term future and demonstrates how to use this for efficient planning and exploration.
Leveraging Communication Topologies Between Learning Agents in Deep Reinforcement Learning
A common technique to improve learning performance in deep reinforcement learning (DRL) and many other machine learning algorithms is to run multiple learning agents in parallel.
InfoBot: Transfer and Exploration via the Information Bottleneck
In new environments, this model can then identify novel subgoals for further exploration, guiding the agent through a sequence of potential decision states and through new regions of the state space.
A Meta-Transfer Objective for Learning to Disentangle Causal Mechanisms
We show that causal structures can be parameterized via continuous variables and learned end-to-end.
Maximum Entropy Generators for Energy-Based Models
Maximum likelihood estimation of energy-based models is a challenging problem due to the intractability of the log-likelihood gradient.
Learning powerful policies and better dynamics models by encouraging consistency
Analogously, we would expect such interaction to be helpful for a learning agent while learning to model the environment dynamics.
Sparse Attentive Backtracking: Temporal CreditAssignment Through Reminding
We consider the hypothesis that such memory associations between past and present could be used for credit assignment through arbitrarily long sequences, propagating the credit assigned to the current state to the associated past state.
Generalization of Equilibrium Propagation to Vector Field Dynamics
The biological plausibility of the backpropagation algorithm has long been doubted by neuroscientists.
Fortified Networks: Improving the Robustness of Deep Networks by Modeling the Manifold of Hidden Representations
Deep networks have achieved impressive results across a variety of important tasks.
Recall Traces: Backtracking Models for Efficient Reinforcement Learning
In many environments only a tiny subset of all states yield high reward.
Extending the Framework of Equilibrium Propagation to General Dynamics
The biological plausibility of the backpropagation algorithm has long been doubted by neuroscientists.
Z-Forcing: Training Stochastic Recurrent Networks
Stochastic recurrent models have been successful in capturing the variability observed in natural sequential data such as speech.
ACtuAL: Actor-Critic Under Adversarial Learning
This makes it fundamentally difficult to train GANs with discrete data, as generation in this case typically involves a non-differentiable function.
Variational Walkback: Learning a Transition Operator as a Stochastic Recurrent Net
The energy function is then modified so the model and data distributions match, with no guarantee on the number of steps required for the Markov chain to converge.
Sparse Attentive Backtracking: Long-Range Credit Assignment in Recurrent Networks
A major drawback of backpropagation through time (BPTT) is the difficulty of learning long-term dependencies, coming from having to propagate credit information backwards through every single step of the forward computation.
Professor Forcing: A New Algorithm for Training Recurrent Networks
We introduce the Professor Forcing algorithm, which uses adversarial domain adaptation to encourage the dynamics of the recurrent network to be the same when training the network and when sampling from the network over multiple time steps.
An Actor-Critic Algorithm for Sequence Prediction
We present an approach to training neural networks to generate sequences using actor-critic methods from reinforcement learning (RL).
Ranked #8 on
Machine Translation
on IWSLT2015 English-German
Zoneout: Regularizing RNNs by Randomly Preserving Hidden Activations
We propose zoneout, a novel method for regularizing RNNs.
Stories in the Eye: Contextual Visual Interactions for Efficient Video to Language Translation
Integrating higher level visual and linguistic interpretations is at the heart of human intelligence.
