What’s the p(Doom)? And what’s my 5-year track record at AI Predictions? Jordan Boyd-Graber 12,7 тыс. подписчиков Скачать
Meta released a Diplomacy-playing LLM. How good is Cicero at talking to players? [Research Talk] Скачать
QA is not one size fits all: Getting different answers to the same question from an AI [Lecture] Скачать
Natural Questions: Google's QA Dataset Five Years Later and Why it's Impossible Today [Lecture] Скачать
Recurrent Neural Networks as Language Models and the two Tricks that Made them Work [Lecture] Скачать
Explaining Recurrent Neural Networks through a silly Word-Counting Sentiment Example [Lecture] Скачать
The Viterbi Algorithm for Finding a Sentence's Parts of Speech with a Hidden Markov Model [Lecture] Скачать
Why should (Computational) Linguists use the Grammar of Graphics packages like ggplot2? [Lecture] Скачать
Course Evaluations: Why they Stink, Why they're Important, How to Make the Most of Them [Lecture] Скачать
How Accessibility Tools (Subs, Audio Descriptions) Help Computers "See" & Answer Questions [Lecture] Скачать
Multihop question answering: following links, text query generation, dense retrieval [Lecture] Скачать
How computers answer "why" or "how" (long) questions? And why is it so hard to evaluate? [Lecture] Скачать
How to train a computer to answer questions in a conversation: Datasets and evaluation [Lecture] Скачать
What happens if you give computers a "closed book" exam: Generative QA models (T5, GPT3) [Lecture] Скачать
[Lecture] Knowledge Bases for Answering Questions: The Bridge from Good Old Fashioned AI to Today Скачать
[Lecture] How did computers answer questions before Siri and Alexa? BASEBALL, LUNAR, and SHRDLU Скачать
[Lecture] How a Tiny College of Aeronautics in Bedfordshire Made the Modern Internet Possible Скачать
[Lecture] How do computers know if a sentence is true? An intro to Natural Language Inference (NLI) Скачать
[Expo] Can Trivia and ML Experts Beat and AI designed by UMD Undergrads? Final Project for CMSC 470 Скачать
Is Automated Topic Model Evaluation Broken?: The Incoherence of Coherence [Paper Read Out Loud] Скачать
What’s in a Name? Answer Equivalence For OpenDomain Question Answering [2021 EMNLP Research Talk] Скачать
Paper Read Aloud: Evaluation Examples Are Not Equally Informative: Should NLP Leaderboards Change? Скачать
[Research] Evaluation Examples Are Not Equally Informative: How Should That Change NLP Leaderboards? Скачать
What kind of a job can you get with a degree in Natural Language Processing (w/ Philip Resnik) Скачать
Where can (and should) you publish your natural language processing research? (w/ Philip Resnik) Скачать
New Course: Quantitative Research Methods (INST 808), Deep Learning for Information Scientists Скачать
Why Overfitting Isn’t Always Bad: Retrofitting Cross-Lingual Word Embeddings to Dictionaries. Скачать
Tandem Anchoring: a Multiword Anchor Approach for Interactive Topic Modeling Jeffrey Lund, Conn Скачать