Search Results for author:
Found 78 papers, 59 papers with code
Bio-xLSTM: Generative modeling, representation and in-context learning of biological and chemical sequences
While Transformers have yielded impressive results, their quadratic runtime dependency on the sequence length complicates their use for long genomic sequences and in-context learning on proteins and chemical sequences.
A Large Recurrent Action Model: xLSTM enables Fast Inference for Robotics Tasks
In recent years, there has been a trend in the field of Reinforcement Learning (RL) towards large action models trained offline on large-scale datasets via sequence modeling.
The Disparate Benefits of Deep Ensembles
Furthermore, we identify the per-group difference in predictive diversity of ensemble members as the potential cause of the disparate benefits effect.
On Information-Theoretic Measures of Predictive Uncertainty
Reliable estimation of predictive uncertainty is crucial for machine learning applications, particularly in high-stakes scenarios where hedging against risks is essential.
Retrieval-Augmented Decision Transformer: External Memory for In-context RL
Furthermore, we illuminate the limitations of current in-context RL methods on complex environments and discuss future directions.
One Initialization to Rule them All: Fine-tuning via Explained Variance Adaptation
Then, we initialize the LoRA matrices with the obtained right-singular vectors and re-distribute ranks among all weight matrices to explain the maximal amount of variance and continue the standard LoRA fine-tuning procedure.
Contrastive Abstraction for Reinforcement Learning
Contrastive abstraction learning first constructs clusters of state representations by contrastive learning and then applies modern Hopfield networks to determine the abstract states.
Simplified priors for Object-Centric Learning
Our method encodes the input image with a Convolutional Neural Network and then uses a branch of alternating Convolution and MaxPool layers to create specialized sub-networks and extract primitive slots.
Comparison Visual Instruction Tuning
Comparing two images in terms of Commonalities and Differences (CaD) is a fundamental human capability that forms the basis of advanced visual reasoning and interpretation.
Vision-LSTM: xLSTM as Generic Vision Backbone
Transformers are widely used as generic backbones in computer vision, despite initially introduced for natural language processing.
Semantically Diverse Language Generation for Uncertainty Estimation in Language Models
Current LLMs generate text in an autoregressive fashion by predicting and appending text tokens.
A Diffusion Model Framework for Unsupervised Neural Combinatorial Optimization
Learning to sample from intractable distributions over discrete sets without relying on corresponding training data is a central problem in a wide range of fields, including Combinatorial Optimization.
Large Language Models Can Self-Improve At Web Agent Tasks
Recent research has also demonstrated LLMs have the capability to exceed their base performance through self-improvement, i. e. fine-tuning on data generated by the model itself.
Energy-based Hopfield Boosting for Out-of-Distribution Detection
Out-of-distribution (OOD) detection is critical when deploying machine learning models in the real world.
Out-of-Distribution Detection
Out of Distribution (OOD) Detection
xLSTM: Extended Long Short-Term Memory
In the 1990s, the constant error carousel and gating were introduced as the central ideas of the Long Short-Term Memory (LSTM).
VN-EGNN: E(3)-Equivariant Graph Neural Networks with Virtual Nodes Enhance Protein Binding Site Identification
However, the performance of GNNs at binding site identification is still limited potentially due to the lack of dedicated nodes that model hidden geometric entities, such as binding pockets.
Geometry-Informed Neural Networks
Geometry is a ubiquitous tool in computer graphics, design, and engineering.
Overcoming Saturation in Density Ratio Estimation by Iterated Regularization
Estimating the ratio of two probability densities from finitely many samples, is a central task in machine learning and statistics.
MIM-Refiner: A Contrastive Learning Boost from Intermediate Pre-Trained Representations
We introduce MIM (Masked Image Modeling)-Refiner, a contrastive learning boost for pre-trained MIM models.
Ranked #2 on
Image Clustering
on ImageNet
SymbolicAI: A framework for logic-based approaches combining generative models and solvers
Through these operations based on in-context learning our framework enables the creation and evaluation of explainable computational graphs.
Variational Annealing on Graphs for Combinatorial Optimization
Several recent unsupervised learning methods use probabilistic approaches to solve combinatorial optimization (CO) problems based on the assumption of statistically independent solution variables.
Introducing an Improved Information-Theoretic Measure of Predictive Uncertainty
Our analyses show that the current measure erroneously assumes that the BMA predictive distribution is equivalent to the predictive distribution of the true model that generated the dataset.
Functional trustworthiness of AI systems by statistically valid testing
The authors are concerned about the safety, health, and rights of the European citizens due to inadequate measures and procedures required by the current draft of the EU Artificial Intelligence (AI) Act for the conformity assessment of AI systems.
Linear Alignment of Vision-language Models for Image Captioning
We propose a more efficient training protocol that fits a linear mapping between image and text embeddings of CLIP via a closed-form solution.
Learning to Modulate pre-trained Models in RL
That is, the performance on the pre-training tasks deteriorates when fine-tuning on new tasks.
Semantic HELM: A Human-Readable Memory for Reinforcement Learning
Then we feed these tokens to a pretrained language model that serves the agent as memory and provides it with a coherent and human-readable representation of the past.
Addressing Parameter Choice Issues in Unsupervised Domain Adaptation by Aggregation
Our method outperforms deep embedded validation (DEV) and importance weighted validation (IWV) on all datasets, setting a new state-of-the-art performance for solving parameter choice issues in unsupervised domain adaptation with theoretical error guarantees.
Context-enriched molecule representations improve few-shot drug discovery
Our novel concept for molecule representation enrichment is to associate molecules from both the support set and the query set with a large set of reference (context) molecules through a Modern Hopfield Network.
Contrastive Tuning: A Little Help to Make Masked Autoencoders Forget
In this work, we study how to combine the efficiency and scalability of MIM with the ability of ID to perform downstream classification in the absence of large amounts of labeled data.
Ranked #1 on
Image Clustering
on Imagenet-dog-15
(using extra training data)
Conformal Prediction for Time Series with Modern Hopfield Networks
To quantify uncertainty, conformal prediction methods are gaining continuously more interest and have already been successfully applied to various domains.
Traffic4cast at NeurIPS 2022 -- Predict Dynamics along Graph Edges from Sparse Node Data: Whole City Traffic and ETA from Stationary Vehicle Detectors
We only provide vehicle count data from spatially sparse stationary vehicle detectors in these three cities as model input for this task.
Enhancing Activity Prediction Models in Drug Discovery with the Ability to Understand Human Language
Activity and property prediction models are the central workhorses in drug discovery and materials sciences, but currently they have to be trained or fine-tuned for new tasks.
G-Signatures: Global Graph Propagation With Randomized Signatures
Graph neural networks (GNNs) have evolved into one of the most popular deep learning architectures.
Txt2Img-MHN: Remote Sensing Image Generation from Text Using Modern Hopfield Networks
To better evaluate the realism and semantic consistency of the generated images, we further conduct zero-shot classification on real remote sensing data using the classification model trained on synthesized images.
Reactive Exploration to Cope with Non-Stationarity in Lifelong Reinforcement Learning
Therefore, exploration strategies and learning methods are required that are capable of tracking the steady domain shifts, and adapting to them.
Few-Shot Learning by Dimensionality Reduction in Gradient Space
We introduce SubGD, a novel few-shot learning method which is based on the recent finding that stochastic gradient descent updates tend to live in a low-dimensional parameter subspace.
Entangled Residual Mappings
Residual mappings have been shown to perform representation learning in the first layers and iterative feature refinement in higher layers.
Hopular: Modern Hopfield Networks for Tabular Data
In experiments on small-sized tabular datasets with less than 1, 000 samples, Hopular surpasses Gradient Boosting, Random Forests, SVMs, and in particular several Deep Learning methods.
Ranked #1 on
General Classification
on Shrutime
History Compression via Language Models in Reinforcement Learning
We propose to utilize a frozen Pretrained Language Transformer (PLT) for history representation and compression to improve sample efficiency.
Traffic4cast at NeurIPS 2021 -- Temporal and Spatial Few-Shot Transfer Learning in Gridded Geo-Spatial Processes
The IARAI Traffic4cast competitions at NeurIPS 2019 and 2020 showed that neural networks can successfully predict future traffic conditions 1 hour into the future on simply aggregated GPS probe data in time and space bins.
Improving Few- and Zero-Shot Reaction Template Prediction Using Modern Hopfield Networks
Finding synthesis routes for molecules of interest is essential in the discovery of new drugs and materials.
Ranked #26 on
Single-step retrosynthesis
on USPTO-50k
A Dataset Perspective on Offline Reinforcement Learning
The dataset characteristics are determined by the behavioral policy that samples this dataset.
CLOOB: Modern Hopfield Networks with InfoLOOB Outperform CLIP
We suggest to use modern Hopfield networks to tackle the problem of explaining away.
Boundary Graph Neural Networks for 3D Simulations
However, it is notoriously difficult to integrate them into machine learning approaches due to their heterogeneity with respect to size and orientation.
Learning 3D Granular Flow Simulations
Recently, the application of machine learning models has gained momentum in natural sciences and engineering, which is a natural fit due to the abundance of data in these fields.
Modern Hopfield Networks for Few- and Zero-Shot Reaction Template Prediction
Finding synthesis routes for molecules of interest is an essential step in the discovery of new drugs and materials.
Trusted Artificial Intelligence: Towards Certification of Machine Learning Applications
Artificial Intelligence is one of the fastest growing technologies of the 21st century and accompanies us in our daily lives when interacting with technical applications.
MC-LSTM: Mass-Conserving LSTM
MC-LSTMs set a new state-of-the-art for neural arithmetic units at learning arithmetic operations, such as addition tasks, which have a strong conservation law, as the sum is constant over time.
Uncertainty Estimation with Deep Learning for Rainfall-Runoff Modelling
Deep Learning is becoming an increasingly important way to produce accurate hydrological predictions across a wide range of spatial and temporal scales.
Convergence Proof for Actor-Critic Methods Applied to PPO and RUDDER
We prove under commonly used assumptions the convergence of actor-critic reinforcement learning algorithms, which simultaneously learn a policy function, the actor, and a value function, the critic.
Rainfall-Runoff Prediction at Multiple Timescales with a Single Long Short-Term Memory Network
Compared to naive prediction with a distinct LSTM per timescale, the multi-timescale architectures are computationally more efficient with no loss in accuracy.
Cross-Domain Few-Shot Learning by Representation Fusion
On the few-shot datasets miniImagenet and tieredImagenet with small domain shifts, CHEF is competitive with state-of-the-art methods.
Align-RUDDER: Learning From Few Demonstrations by Reward Redistribution
For such complex tasks, the recently proposed RUDDER uses reward redistribution to leverage steps in the Q-function that are associated with accomplishing sub-tasks.
Hopfield Networks is All You Need
The new update rule is equivalent to the attention mechanism used in transformers.
Immune Repertoire Classification
Multiple Instance Learning
+1
Modern Hopfield Networks and Attention for Immune Repertoire Classification
We show that the attention mechanism of transformer architectures is actually the update rule of modern Hopfield networks that can store exponentially many patterns.
Large-scale ligand-based virtual screening for SARS-CoV-2 inhibitors using deep neural networks
Due to the current severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic, there is an urgent need for novel therapies and drugs.
Detecting cutaneous basal cell carcinomas in ultra-high resolution and weakly labelled histopathological images
The aim of this study is to evaluate whether it is possible to detect basal cell carcinomas in histological sections using attention-based deep learning models and to overcome the ultra-high resolution and the weak labels of whole slide images.
Using LSTMs for climate change assessment studies on droughts and floods
Climate change affects occurrences of floods and droughts worldwide.
Quantum Optical Experiments Modeled by Long Short-Term Memory
In this work, we show that machine learning models can provide significant improvement over random search.
Patch Refinement -- Localized 3D Object Detection
We introduce Patch Refinement a two-stage model for accurate 3D object detection and localization from point cloud data.
Ranked #1 on
Object Detection
on KITTI Cars Moderate
Explaining and Interpreting LSTMs
While neural networks have acted as a strong unifying force in the design of modern AI systems, the neural network architectures themselves remain highly heterogeneous due to the variety of tasks to be solved.
A GAN based solver of black-box inverse problems
We propose a GAN based approach to solve inverse problems which have non-differential or non-continuous forward relations.
Towards Learning Universal, Regional, and Local Hydrological Behaviors via Machine-Learning Applied to Large-Sample Datasets
The problem currently is that traditional hydrological models degrade significantly in performance when calibrated for multiple basins together instead of for a single basin alone.
Human-level Protein Localization with Convolutional Neural Networks
We present the largest comparison of CNN architectures including GapNet-PL for protein localization in HTI images of human cells.
End-to-end learning of pharmacological assays from high-resolution microscopy images
Surprisingly, we could predict 29% of the 209 pharmacological assays at high predictive performance (AUC > 0. 9).
NeuralHydrology -- Interpreting LSTMs in Hydrology
LSTMs are particularly well-suited for this problem since memory cells can represent dynamic reservoirs and storages, which are essential components in state-space modelling approaches of the hydrological system.
Interpretable Deep Learning in Drug Discovery
Without any means of interpretation, neural networks that predict molecular properties and bioactivities are merely black boxes.
RUDDER: Return Decomposition for Delayed Rewards
In MDPs the Q-values are equal to the expected immediate reward plus the expected future rewards.
Ranked #9 on
Atari Games
on Atari 2600 Bowling
Fréchet ChemNet Distance: A metric for generative models for molecules in drug discovery
We propose a novel distance measure between two sets of molecules, called Fr\'echet ChemNet distance (FCD), that can be used as an evaluation metric for generative models.
First Order Generative Adversarial Networks
To formally describe an optimal update direction, we introduce a theoretical framework which allows the derivation of requirements on both the divergence and corresponding method for determining an update direction, with these requirements guaranteeing unbiased mini-batch updates in the direction of steepest descent.
Ranked #3 on
Image Generation
on LSUN Bedroom 64 x 64
DeepSynergy: Predicting Anti-Cancer Drug Synergy with Deep Learning
While drug combination therapies are a well-established concept in cancer treatment, identifying novel synergistic combinations is challenging due to the size of combinatorial space.
Coulomb GANs: Provably Optimal Nash Equilibria via Potential Fields
We prove that Coulomb GANs possess only one Nash equilibrium which is optimal in the sense that the model distribution equals the target distribution.
GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
Generative Adversarial Networks (GANs) excel at creating realistic images with complex models for which maximum likelihood is infeasible.
Ranked #2 on
Image Generation
on LSUN Bedroom 64 x 64
Self-Normalizing Neural Networks
We introduce self-normalizing neural networks (SNNs) to enable high-level abstract representations.
Ranked #8 on
Drug Discovery
on Tox21
Speeding up Semantic Segmentation for Autonomous Driving
We propose a novel deep network architecture for image segmentation that keeps the high accuracy while being efficient enough for embedded devices.
Fast and Accurate Deep Network Learning by Exponential Linear Units (ELUs)
In contrast to ReLUs, ELUs have negative values which allows them to push mean unit activations closer to zero like batch normalization but with lower computational complexity.
Ranked #143 on
Image Classification
on CIFAR-100
(using extra training data)
Toxicity Prediction using Deep Learning
The goal of this challenge was to assess the performance of computational methods in predicting the toxicity of chemical compounds.
