Niket Tandon
language models
generation
transformers
commonsense
reasoning
information retrieval
fact-checking
large language models
interactive nlp
human in the loop
entity
gpt
data generation
interactive machine learning
gpt3
8
presentations
5
number of views
SHORT BIO
Niket Tandon is a senior research scientist at the Allen Institute for AI in Seattle. He is broadly interested in NLP and AI, with a focus on injecting commonsense into deep learning models at the Aristo team that created AI which aced science exams. He completed his PhD from the Max Planck Institute for Informatics in Germany in 2016, resulting in the largest automatically extracted commonsense knowledge base called WebChild. He is the founder of PQRS research, an organization that provides a research footing to undergrads from underrepresented institutes.
Presentations
Calibrating Large Language Models with Sample Consistency
Qing Lyu and 8 other authors
Let Me Teach You: Pedagogical Foundations of Feedback for Language Models
Beatriz Borges and 3 other authors
Tailoring with Targeted Precision: Edit-Based Agents for Open-Domain Procedure Customization
Yash Kumar Lal and 5 other authors
OpenPI2.0: An Improved Dataset for Entity Tracking in Texts
Li Zhang and 4 other authors
Editing Common Sense in Transformers | VIDEO
Anshita Gupta and 6 other authors
Learning to repair: Repairing model output errors after deployment using a dynamic memory of feedback
Niket Tandon and 3 other authors
GUD-IR: Generative Retrieval for Semiparametric Models
Niket Tandon and 3 other authors
Memory-assisted prompt editing to improve GPT-3 after deployment
Niket Tandon and 3 other authors