Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Roger Wattenhofer

Roger Wattenhofer

On Isotropy Calibration of Transformers


Sep 27, 2021
Yue Ding, Karolis Martinkus, Damian Pascual, Simon Clematide, Roger Wattenhofer

Figure 1 for On Isotropy Calibration of Transformers
Figure 2 for On Isotropy Calibration of Transformers
Figure 3 for On Isotropy Calibration of Transformers
Figure 4 for On Isotropy Calibration of Transformers

  Access Paper or Ask Questions

A Plug-and-Play Method for Controlled Text Generation


Sep 20, 2021
Damian Pascual, Beni Egressy, Clara Meister, Ryan Cotterell, Roger Wattenhofer

Figure 1 for A Plug-and-Play Method for Controlled Text Generation
Figure 2 for A Plug-and-Play Method for Controlled Text Generation
Figure 3 for A Plug-and-Play Method for Controlled Text Generation
Figure 4 for A Plug-and-Play Method for Controlled Text Generation
* Findings of EMNLP 2021 
  Access Paper or Ask Questions

BERT is Robust! A Case Against Synonym-Based Adversarial Examples in Text Classification


Sep 15, 2021
Jens Hauser, Zhao Meng, Damián Pascual, Roger Wattenhofer

Figure 1 for BERT is Robust! A Case Against Synonym-Based Adversarial Examples in Text Classification
Figure 2 for BERT is Robust! A Case Against Synonym-Based Adversarial Examples in Text Classification
Figure 3 for BERT is Robust! A Case Against Synonym-Based Adversarial Examples in Text Classification
Figure 4 for BERT is Robust! A Case Against Synonym-Based Adversarial Examples in Text Classification
* 12 pages with appendix, 7 figures 
  Access Paper or Ask Questions

Self-Supervised Contrastive Learning with Adversarial Perturbations for Robust Pretrained Language Models


Jul 15, 2021
Zhao Meng, Yihan Dong, Mrinmaya Sachan, Roger Wattenhofer

Figure 1 for Self-Supervised Contrastive Learning with Adversarial Perturbations for Robust Pretrained Language Models
Figure 2 for Self-Supervised Contrastive Learning with Adversarial Perturbations for Robust Pretrained Language Models
Figure 3 for Self-Supervised Contrastive Learning with Adversarial Perturbations for Robust Pretrained Language Models
Figure 4 for Self-Supervised Contrastive Learning with Adversarial Perturbations for Robust Pretrained Language Models
* Work in progress 
  Access Paper or Ask Questions

Towards BERT-based Automatic ICD Coding: Limitations and Opportunities


Apr 14, 2021
Damian Pascual, Sandro Luck, Roger Wattenhofer

Figure 1 for Towards BERT-based Automatic ICD Coding: Limitations and Opportunities
Figure 2 for Towards BERT-based Automatic ICD Coding: Limitations and Opportunities
Figure 3 for Towards BERT-based Automatic ICD Coding: Limitations and Opportunities
Figure 4 for Towards BERT-based Automatic ICD Coding: Limitations and Opportunities
* Accepted at BioNLP 2021 
  Access Paper or Ask Questions

Should Graph Neural Networks Use Features, Edges, Or Both?


Mar 11, 2021
Lukas Faber, Yifan Lu, Roger Wattenhofer

Figure 1 for Should Graph Neural Networks Use Features, Edges, Or Both?
Figure 2 for Should Graph Neural Networks Use Features, Edges, Or Both?
Figure 3 for Should Graph Neural Networks Use Features, Edges, Or Both?
Figure 4 for Should Graph Neural Networks Use Features, Edges, Or Both?

  Access Paper or Ask Questions

Towards Robust Graph Contrastive Learning


Feb 25, 2021
Nikola Jovanović, Zhao Meng, Lukas Faber, Roger Wattenhofer

Figure 1 for Towards Robust Graph Contrastive Learning
Figure 2 for Towards Robust Graph Contrastive Learning
Figure 3 for Towards Robust Graph Contrastive Learning
Figure 4 for Towards Robust Graph Contrastive Learning

  Access Paper or Ask Questions

Of Non-Linearity and Commutativity in BERT


Jan 14, 2021
Sumu Zhao, Damian Pascual, Gino Brunner, Roger Wattenhofer

Figure 1 for Of Non-Linearity and Commutativity in BERT
Figure 2 for Of Non-Linearity and Commutativity in BERT
Figure 3 for Of Non-Linearity and Commutativity in BERT
Figure 4 for Of Non-Linearity and Commutativity in BERT

  Access Paper or Ask Questions

KM-BART: Knowledge Enhanced Multimodal BART for Visual Commonsense Generation


Jan 02, 2021
Yiran Xing, Zai Shi, Zhao Meng, Yunpu Ma, Roger Wattenhofer

Figure 1 for KM-BART: Knowledge Enhanced Multimodal BART for Visual Commonsense Generation
Figure 2 for KM-BART: Knowledge Enhanced Multimodal BART for Visual Commonsense Generation
Figure 3 for KM-BART: Knowledge Enhanced Multimodal BART for Visual Commonsense Generation
Figure 4 for KM-BART: Knowledge Enhanced Multimodal BART for Visual Commonsense Generation
* Work in progress. The first three authors contribute equally to this work 
  Access Paper or Ask Questions

Directed Beam Search: Plug-and-Play Lexically Constrained Language Generation


Dec 31, 2020
Damian Pascual, Beni Egressy, Florian Bolli, Roger Wattenhofer

Figure 1 for Directed Beam Search: Plug-and-Play Lexically Constrained Language Generation
Figure 2 for Directed Beam Search: Plug-and-Play Lexically Constrained Language Generation
Figure 3 for Directed Beam Search: Plug-and-Play Lexically Constrained Language Generation
Figure 4 for Directed Beam Search: Plug-and-Play Lexically Constrained Language Generation
* Preprint. Work in progress 
  Access Paper or Ask Questions

Contrastive Graph Neural Network Explanation


Oct 26, 2020
Lukas Faber, Amin K. Moghaddam, Roger Wattenhofer

Figure 1 for Contrastive Graph Neural Network Explanation
Figure 2 for Contrastive Graph Neural Network Explanation
Figure 3 for Contrastive Graph Neural Network Explanation
Figure 4 for Contrastive Graph Neural Network Explanation
* ICML 2020 Workshop on Graph Representation Learning and Beyond (GRL+) 
  Access Paper or Ask Questions

Brain2Word: Decoding Brain Activity for Language Generation


Oct 13, 2020
Nicolas Affolter, Beni Egressy, Damian Pascual, Roger Wattenhofer

Figure 1 for Brain2Word: Decoding Brain Activity for Language Generation
Figure 2 for Brain2Word: Decoding Brain Activity for Language Generation
Figure 3 for Brain2Word: Decoding Brain Activity for Language Generation
Figure 4 for Brain2Word: Decoding Brain Activity for Language Generation

  Access Paper or Ask Questions

A Geometry-Inspired Attack for Generating Natural Language Adversarial Examples


Oct 03, 2020
Zhao Meng, Roger Wattenhofer

Figure 1 for A Geometry-Inspired Attack for Generating Natural Language Adversarial Examples
Figure 2 for A Geometry-Inspired Attack for Generating Natural Language Adversarial Examples
Figure 3 for A Geometry-Inspired Attack for Generating Natural Language Adversarial Examples
Figure 4 for A Geometry-Inspired Attack for Generating Natural Language Adversarial Examples
* COLING 2020 Long Paper 
  Access Paper or Ask Questions

Medley2K: A Dataset of Medley Transitions


Aug 25, 2020
Lukas Faber, Sandro Luck, Damian Pascual, Andreas Roth, Gino Brunner, Roger Wattenhofer

Figure 1 for Medley2K: A Dataset of Medley Transitions
Figure 2 for Medley2K: A Dataset of Medley Transitions
* MML 2020 - 13th Int. Workshop on Machine Learning and Music at ECML-PKDD 2020 
  Access Paper or Ask Questions

Normalized Attention Without Probability Cage


May 19, 2020
Oliver Richter, Roger Wattenhofer

Figure 1 for Normalized Attention Without Probability Cage
Figure 2 for Normalized Attention Without Probability Cage
Figure 3 for Normalized Attention Without Probability Cage
Figure 4 for Normalized Attention Without Probability Cage
* Preprint, work in progress. Feedback welcome 
  Access Paper or Ask Questions

Neural Status Registers


Apr 15, 2020
Lukas Faber, Roger Wattenhofer

Figure 1 for Neural Status Registers
Figure 2 for Neural Status Registers
Figure 3 for Neural Status Registers
Figure 4 for Neural Status Registers

  Access Paper or Ask Questions

Telling BERT's full story: from Local Attention to Global Aggregation


Apr 10, 2020
Damian Pascual, Gino Brunner, Roger Wattenhofer

Figure 1 for Telling BERT's full story: from Local Attention to Global Aggregation
Figure 2 for Telling BERT's full story: from Local Attention to Global Aggregation
Figure 3 for Telling BERT's full story: from Local Attention to Global Aggregation
Figure 4 for Telling BERT's full story: from Local Attention to Global Aggregation
* Preprint. Work in progress 
  Access Paper or Ask Questions

On the Validity of Self-Attention as Explanation in Transformer Models


Aug 12, 2019
Gino Brunner, Yang Liu, Damián Pascual, Oliver Richter, Roger Wattenhofer

Figure 1 for On the Validity of Self-Attention as Explanation in Transformer Models
Figure 2 for On the Validity of Self-Attention as Explanation in Transformer Models
Figure 3 for On the Validity of Self-Attention as Explanation in Transformer Models
Figure 4 for On the Validity of Self-Attention as Explanation in Transformer Models
* Preprint. Work in progress 
  Access Paper or Ask Questions

Synthetic Epileptic Brain Activities Using Generative Adversarial Networks


Jul 22, 2019
Damian Pascual, Amir Aminifar, David Atienza, Philippe Ryvlin, Roger Wattenhofer

Figure 1 for Synthetic Epileptic Brain Activities Using Generative Adversarial Networks
Figure 2 for Synthetic Epileptic Brain Activities Using Generative Adversarial Networks
Figure 3 for Synthetic Epileptic Brain Activities Using Generative Adversarial Networks
Figure 4 for Synthetic Epileptic Brain Activities Using Generative Adversarial Networks

  Access Paper or Ask Questions

Attentive Multi-Task Deep Reinforcement Learning


Jul 05, 2019
Timo Bram, Gino Brunner, Oliver Richter, Roger Wattenhofer

Figure 1 for Attentive Multi-Task Deep Reinforcement Learning
Figure 2 for Attentive Multi-Task Deep Reinforcement Learning
Figure 3 for Attentive Multi-Task Deep Reinforcement Learning
Figure 4 for Attentive Multi-Task Deep Reinforcement Learning
* Accepted as conference paper at ECML PKDD 2019 
  Access Paper or Ask Questions

Quantile Regression Deep Reinforcement Learning


Jun 27, 2019
Oliver Richter, Roger Wattenhofer

Figure 1 for Quantile Regression Deep Reinforcement Learning
Figure 2 for Quantile Regression Deep Reinforcement Learning
Figure 3 for Quantile Regression Deep Reinforcement Learning
Figure 4 for Quantile Regression Deep Reinforcement Learning
* Preprint 
  Access Paper or Ask Questions

Using State Predictions for Value Regularization in Curiosity Driven Deep Reinforcement Learning


Sep 30, 2018
Gino Brunner, Manuel Fritsche, Oliver Richter, Roger Wattenhofer

Figure 1 for Using State Predictions for Value Regularization in Curiosity Driven Deep Reinforcement Learning
Figure 2 for Using State Predictions for Value Regularization in Curiosity Driven Deep Reinforcement Learning
Figure 3 for Using State Predictions for Value Regularization in Curiosity Driven Deep Reinforcement Learning
Figure 4 for Using State Predictions for Value Regularization in Curiosity Driven Deep Reinforcement Learning

  Access Paper or Ask Questions

The Urban Last Mile Problem: Autonomous Drone Delivery to Your Balcony


Sep 21, 2018
Gino Brunner, Bence Szebedy, Simon Tanner, Roger Wattenhofer

Figure 1 for The Urban Last Mile Problem: Autonomous Drone Delivery to Your Balcony
Figure 2 for The Urban Last Mile Problem: Autonomous Drone Delivery to Your Balcony
Figure 3 for The Urban Last Mile Problem: Autonomous Drone Delivery to Your Balcony
Figure 4 for The Urban Last Mile Problem: Autonomous Drone Delivery to Your Balcony

  Access Paper or Ask Questions

MIDI-VAE: Modeling Dynamics and Instrumentation of Music with Applications to Style Transfer


Sep 20, 2018
Gino Brunner, Andres Konrad, Yuyi Wang, Roger Wattenhofer

Figure 1 for MIDI-VAE: Modeling Dynamics and Instrumentation of Music with Applications to Style Transfer
Figure 2 for MIDI-VAE: Modeling Dynamics and Instrumentation of Music with Applications to Style Transfer
Figure 3 for MIDI-VAE: Modeling Dynamics and Instrumentation of Music with Applications to Style Transfer
Figure 4 for MIDI-VAE: Modeling Dynamics and Instrumentation of Music with Applications to Style Transfer
* Paper accepted at the 19th International Society for Music Information Retrieval Conference, ISMIR 2018, Paris, France 
  Access Paper or Ask Questions

Symbolic Music Genre Transfer with CycleGAN


Sep 20, 2018
Gino Brunner, Yuyi Wang, Roger Wattenhofer, Sumu Zhao

Figure 1 for Symbolic Music Genre Transfer with CycleGAN
Figure 2 for Symbolic Music Genre Transfer with CycleGAN
Figure 3 for Symbolic Music Genre Transfer with CycleGAN
Figure 4 for Symbolic Music Genre Transfer with CycleGAN
* Paper accepted at the 30th International Conference on Tools with Artificial Intelligence, ICTAI 2018, Volos, Greece 
  Access Paper or Ask Questions

Natural Language Multitasking: Analyzing and Improving Syntactic Saliency of Hidden Representations


Jan 18, 2018
Gino Brunner, Yuyi Wang, Roger Wattenhofer, Michael Weigelt

Figure 1 for Natural Language Multitasking: Analyzing and Improving Syntactic Saliency of Hidden Representations
Figure 2 for Natural Language Multitasking: Analyzing and Improving Syntactic Saliency of Hidden Representations
Figure 3 for Natural Language Multitasking: Analyzing and Improving Syntactic Saliency of Hidden Representations
Figure 4 for Natural Language Multitasking: Analyzing and Improving Syntactic Saliency of Hidden Representations
* The 31st Annual Conference on Neural Information Processing (NIPS) - Workshop on Learning Disentangled Features: from Perception to Control, Long Beach, CA, December 2017 
  Access Paper or Ask Questions

JamBot: Music Theory Aware Chord Based Generation of Polyphonic Music with LSTMs


Nov 21, 2017
Gino Brunner, Yuyi Wang, Roger Wattenhofer, Jonas Wiesendanger

Figure 1 for JamBot: Music Theory Aware Chord Based Generation of Polyphonic Music with LSTMs
Figure 2 for JamBot: Music Theory Aware Chord Based Generation of Polyphonic Music with LSTMs
Figure 3 for JamBot: Music Theory Aware Chord Based Generation of Polyphonic Music with LSTMs
Figure 4 for JamBot: Music Theory Aware Chord Based Generation of Polyphonic Music with LSTMs
* Paper presented at the 29th International Conference on Tools with Artificial Intelligence, ICTAI 2017, Boston, MA, USA 
  Access Paper or Ask Questions

Teaching a Machine to Read Maps with Deep Reinforcement Learning


Nov 20, 2017
Gino Brunner, Oliver Richter, Yuyi Wang, Roger Wattenhofer

Figure 1 for Teaching a Machine to Read Maps with Deep Reinforcement Learning
Figure 2 for Teaching a Machine to Read Maps with Deep Reinforcement Learning
Figure 3 for Teaching a Machine to Read Maps with Deep Reinforcement Learning
Figure 4 for Teaching a Machine to Read Maps with Deep Reinforcement Learning
* Paper accepted at 32nd AAAI Conference on Artificial Intelligence, AAAI 2018, New Orleans, Louisiana, USA 
  Access Paper or Ask Questions