Name: Dennis Aumiller
Type: User
Company: Heidelberg University
Bio: CS PhD student at Heidelberg University, working on Text Summarization. Python is awesome, and so is Huggingface
Twitter: d_aumiller
Location: Heidelberg, Germany
Blog: https://dennis-aumiller.de
Dennis Aumiller's Projects
Official style files for papers submitted to venues of the Association for Computational Linguistics
A simple tool that allows you to annotate tokens in a sentence via a React interface
Details about BootCI module which implements bootstrap confidence intervals (CIs) and Fisher-Pitman test for paired samples
Python Library for Accessing the Cohere API
Evaluation script for CQE framework
A small suite of (to me) interesting use cases and comparative implementations in cython.
Dagger - An implementation of Dataset Aggregation
Explaining the concept of Git
Interactive neuroscience tutorial app using Muse and React Native to teach EEG and BCI basics.
Submission for the 2018 UofT BCB BioHacks Challenge
Experimenting with variations of CNNs, specifically with growing and shrinking receptive fields
Code accompanying SSDBM 2020 submission "A Versatile Hypergraph Model for Document Collections" by Spitz et al.
Just another DAgger algorithm implementation
Klexikon: A German Dataset for Joint Summarization and Simplification
lamtram: A toolkit for neural language and translation modeling
⚡ Building applications with LLMs through composability ⚡
Legal Reference Extraction
LLM inference with HuggingFace (experimental)
LongSumm - Scientific Document Summarization Task
Python PDF Parser -- fork with Python 2+3 support using six
Transformation script that aggregates all the PoS annotations in the Penn Treebank available on Kaggle.
My current personal website, geared towards both academic and industry. Implemented on top of hugo-toha
A collection of helpful patterns and comparisons in Python (3.6+)
Updated version of tempeval 3 toolkit for Python3
Scientific Document Summarization Corpus and Annotations from the WING NUS group.
Sentence Embeddings with BERT & XLNet