Aditi Mishra

I am a Computer Science PhD student at School of Computing and Augmented Intelligence, Arizona State University, where I work with Dr. Chris Bryan on data visualization.

My interests particularly lie in designing and developing interactive interfaces to enable autonomous agent interpretability and explanability for novice end users. I am also interested in applying machine learning to reconstruct large unstructured scientific datasets.

Email  /  CV  /  Google Scholar  /  Twitter  /  Linkedin  /  Github

[Jan 2024] I have started as a Data Science Intern at Stanford Healthcare, Palo Alto. I am working on leveraging LLMs to summarize patient health records and ways to evaluate generated summaries.

[May 2023] I have started as a Summer Research Intern at Megagon Labs, Mountain View. I am working on leveraging LLMs to rationalize knowledge augmented model's decisions.

[Sept 2022] I have been awarded the Graduate Travel Fellowship Award by ASU.

[March 2022] I have been awarded the SCAI Doctoral Fellowship by ASU.

[Feb 2022] Two of my works - Why? Why not? When? Visual Explanations of Agent Behavior in Reinforcement Learning and News Kaleidoscope: Visual Investigation of Coverage Diversity in News Event Reporting has been accepted in this years PacificVis held in Tsukuba, Japan.

[May 2020] I started my summer internship at Los Alamos National Lab as a research intern.

Visual analytic system to interactively explore, understand and summarize the difference in Vision Language Model patterns for real and fake images.

Generating and evaluating the proficiency of LLMs in generating rationales grounded on external knowledge for knowledge intensive tasks.

Providing a visual analytics interface to aid easier prompting alterations for LMs based on linguistic and contextual recommendations.

Explaining deep neural networks using image based concepts extracted from ACE+TCAV.

Providing a visual analytics interface to question and thus gain trust in an autonomous agent's decision.

A visual analytic system for journalism expert users to be able to identify underlying polarities in different news organizations.

ChartStory is meant for crafting data stories from a collection of user-created charts.

A study to investigate perception in text-embellished narrative visualizations under Observation, Search and Recall tasks.

An interactive visualization system for the analysis of the VAST 2020 Mini-Challenge 2 (MC2) dataset.

Deep Learning method to reconstruct a large unstructured scientific dataset.