Search Results for author: Ann Yuan

Found 14 papers, 2 papers with code

The Case for a Single Model that can Both Generate Continuations and Fill-in-the-Blank

no code implementations Findings (NAACL) 2022 Daphne Ippolito, Liam Dugan, Emily Reif, Ann Yuan, Andy Coenen, Chris Callison-Burch

While previous work has tackled this problem with models trained specifically to do fill in the blank, a more useful model is one that can effectively perform _both_ FitB and continuation tasks.

Position Text Generation

Who's asking? User personas and the mechanics of latent misalignment

no code implementations17 Jun 2024 Asma Ghandeharioun, Ann Yuan, Marius Guerard, Emily Reif, Michael A. Lepori, Lucas Dixon

In fact, we find manipulating user persona to be even more effective for eliciting harmful content than direct attempts to control model refusal.

ConstitutionalExperts: Training a Mixture of Principle-based Prompts

no code implementations7 Mar 2024 Savvas Petridis, Ben Wedin, Ann Yuan, James Wexler, Nithum Thain

We also show that we can improve overall performance by learning unique prompts for different semantic regions of the training data and using a mixture-of-experts (MoE) architecture to route inputs at inference time.

Gradient-Based Automated Iterative Recovery for Parameter-Efficient Tuning

no code implementations13 Feb 2023 Maximilian Mozes, Tolga Bolukbasi, Ann Yuan, Frederick Liu, Nithum Thain, Lucas Dixon

In this paper, we explore the use of TracIn to improve model performance in the parameter-efficient tuning (PET) setting.

Decision Making Transfer Learning

Towards Agile Text Classifiers for Everyone

no code implementations13 Feb 2023 Maximilian Mozes, Jessica Hoffmann, Katrin Tomanek, Muhamed Kouate, Nithum Thain, Ann Yuan, Tolga Bolukbasi, Lucas Dixon

Text-based safety classifiers are widely used for content moderation and increasingly to tune generative language model behavior - a topic of growing concern for the safety of digital assistants and chatbots.

Language Modelling text-classification +1

Creative Writing with an AI-Powered Writing Assistant: Perspectives from Professional Writers

no code implementations9 Nov 2022 Daphne Ippolito, Ann Yuan, Andy Coenen, Sehmon Burnam

Recent developments in natural language generation (NLG) using neural language models have brought us closer than ever to the goal of building AI-powered creative writing tools.

Text Generation

The Case for a Single Model that can Both Generate Continuations and Fill in the Blank

no code implementations9 Jun 2022 Daphne Ippolito, Liam Dugan, Emily Reif, Ann Yuan, Andy Coenen, Chris Callison-Burch

The task of inserting text into a specified position in a passage, known as fill in the blank (FitB), is useful for a variety of applications where writers interact with a natural language generation (NLG) system to craft text.

Position Text Generation

SynthBio: A Case Study in Human-AI Collaborative Curation of Text Datasets

no code implementations11 Nov 2021 Ann Yuan, Daphne Ippolito, Vitaly Nikolaev, Chris Callison-Burch, Andy Coenen, Sebastian Gehrmann

We use our method to curate SynthBio - a new evaluation set for WikiBio - composed of structured attribute lists describing fictional individuals, mapped to natural language biographies.

Attribute Language Modelling +2

Wordcraft: a Human-AI Collaborative Editor for Story Writing

no code implementations15 Jul 2021 Andy Coenen, Luke Davis, Daphne Ippolito, Emily Reif, Ann Yuan

As neural language models grow in effectiveness, they are increasingly being applied in real-world settings.

Few-Shot Learning

An Interpretability Illusion for BERT

no code implementations14 Apr 2021 Tolga Bolukbasi, Adam Pearce, Ann Yuan, Andy Coenen, Emily Reif, Fernanda Viégas, Martin Wattenberg

We describe an "interpretability illusion" that arises when analyzing the BERT model.

Cannot find the paper you are looking for? You can Submit a new open access paper.