User profiles for H. Strobelt
Hendrik StrobeltSenior Research Scientist IBM Research / MIT-IBM Watson AI Lab Verified email at strobelt.com Cited by 9159 |
Bloom: A 176b-parameter open-access multilingual language model
Large language models (LLMs) have been shown to be able to perform new tasks based on
a few demonstrations or natural language instructions. While these capabilities have led to …
a few demonstrations or natural language instructions. While these capabilities have led to …
UpSet: visualization of intersecting sets
Understanding relationships between sets is an important analysis task that has received
widespread attention in the visualization community. The major challenge in this context is the …
widespread attention in the visualization community. The major challenge in this context is the …
[HTML][HTML] HiGlass: web-based visual exploration and analysis of genome interaction maps
We present HiGlass, an open source visualization tool built on web technologies that
provides a rich interface for rapid, multiplex, and multiscale navigation of 2D genomic maps …
provides a rich interface for rapid, multiplex, and multiscale navigation of 2D genomic maps …
Understanding the role of individual units in a deep neural network
Deep neural networks excel at finding hierarchical representations that solve complex tasks
over large datasets. How can we humans understand these learned representations? In this …
over large datasets. How can we humans understand these learned representations? In this …
Lstmvis: A tool for visual analysis of hidden state dynamics in recurrent neural networks
Recurrent neural networks, and in particular long short-term memory (LSTM) networks, are
a remarkably effective tool for sequence modeling that learn a dense black-box hidden …
a remarkably effective tool for sequence modeling that learn a dense black-box hidden …
Gan dissection: Visualizing and understanding generative adversarial networks
… Recall that ru,P is the one-channel h × w featuremap of unit u in a convolutional generator,
where h × w is typically smaller than the image size. We want to know if a specific unit ru,P …
where h × w is typically smaller than the image size. We want to know if a specific unit ru,P …
[HTML][HTML] Accelerated antimicrobial discovery via deep generative models and molecular dynamics simulations
The de novo design of antimicrobial therapeutics involves the exploration of a vast chemical
repertoire to find compounds with broad-spectrum potency and low toxicity. Here, we report …
repertoire to find compounds with broad-spectrum potency and low toxicity. Here, we report …
Interactive and visual prompt engineering for ad-hoc task adaptation with large language models
… Song, and H. Qu. Understanding hidden memories of recurrent neural networks. … Schick
and H. Schütze. It’s not just size that matters: Small language models are also few-shot learners. …
and H. Schütze. It’s not just size that matters: Small language models are also few-shot learners. …
Extraction of organic chemistry grammar from unsupervised learning of chemical reactions
… H. Strobelt is a visiting research scientist at MIT. Funding: This work was supported by IBM
Research. Author contributions: The project was conceived and planned by PS and BH and …
Research. Author contributions: The project was conceived and planned by PS and BH and …
Gltr: Statistical detection and visualization of generated text
The rapid improvement of language models has raised the specter of abuse of text generation
systems. This progress motivates the development of simple methods for detecting …
systems. This progress motivates the development of simple methods for detecting …