Nikhil Gupta

Nikhil Gupta

San Francisco Bay Area
30K followers 500+ connections

Join to follow

About

Five years ago, I took a leap of faith and started my own company, turning down an Amazon…

Articles by Nikhil

Contributions

Activity

Join now to see all activity

Experience

  • LimeChat Graphic

    LimeChat

    Bengaluru, Karnataka, India

  • -

    New Delhi Area, India

  • -

    Gurgaon, India

  • -

  • -

    Pittsburgh, Pennsilvania

  • -

    Hauz Khas, New Delhi

  • -

    Haus Khas, New Delhi

  • -

    Haus Khas, New Delhi

  • -

    Lausanne Area, Switzerland

  • -

    IIT Delhi

  • -

    Haus Khas, New Delhi

  • -

    New Delhi Area, India

Education

  • Indian Institute of Technology, Delhi Graphic

    Indian Institute of Technology, Delhi

    -

    Activities and Societies: Computer Science Conveyner, Dance Club Representative, Music Club, Robotics Club, Swimming Team

  • -

    Activities and Societies: Environment Society Student Head, Life Skills Society Student Head

  • -

    Activities and Societies: Football Captian, Head Boy, Class Topper

    I was the head boy of my school where I had to take charge of all the school activities and was a link between the administration and the students. I also was the football captain of my house and was awarded as the school topper of my class.

Publications

  • Unsupervised Learning of KB Queries in Task-Oriented Dialogs

    TACL 2021

    Task-oriented dialog (TOD) systems often need to formulate knowledge base (KB) queries corresponding to the user intent and use the query results to generate system responses. Existing approaches require dialog datasets to explicitly annotate these KB queries—these annotations can be time consuming, and expensive. In response, we define the novel problems of predicting the KB query and training the dialog agent, without explicit KB query annotation. For query prediction, we propose a…

    Task-oriented dialog (TOD) systems often need to formulate knowledge base (KB) queries corresponding to the user intent and use the query results to generate system responses. Existing approaches require dialog datasets to explicitly annotate these KB queries—these annotations can be time consuming, and expensive. In response, we define the novel problems of predicting the KB query and training the dialog agent, without explicit KB query annotation. For query prediction, we propose a reinforcement learning (RL) baseline, which rewards the generation of those queries whose KB results cover the entities mentioned in subsequent dialog. Further analysis reveals that correlation among query attributes in KB can significantly confuse memory augmented policy optimization (MAPO), an existing state of the art RL agent.

    To address this, we improve the MAPO baseline with simple but important modifications suited to our task. To train the full TOD system for our setting, we propose a pipelined approach: it independently predicts when to make a KB query (query position predictor), then predicts a KB query at the predicted position (query predictor), and uses the results of predicted query in subsequent dialog (next response predictor). Overall, our work proposes first solutions to our novel problem, and our analysis highlights the research challenges in training TOD systems without query annotation.

    See publication

  • Disentangling Language and Knowledge in Task-Oriented Dialogs

    NAACL 2019

    The Knowledge Base (KB) used for real-world applications, such as booking a movie or restaurant reservation, keeps changing over time. End-to-end neural networks trained for these task-oriented dialogs are expected to be immune to any changes in the KB. However, existing approaches breakdown when asked to handle such changes. We propose an encoder-decoder architecture (BoSsNet) with a novel Bag-of-Sequences (BoSs) memory, which facilitates the disentangled learning of the response's language…

    The Knowledge Base (KB) used for real-world applications, such as booking a movie or restaurant reservation, keeps changing over time. End-to-end neural networks trained for these task-oriented dialogs are expected to be immune to any changes in the KB. However, existing approaches breakdown when asked to handle such changes. We propose an encoder-decoder architecture (BoSsNet) with a novel Bag-of-Sequences (BoSs) memory, which facilitates the disentangled learning of the response's language model and its knowledge incorporation. Consequently, the KB can be modified with new knowledge without a drop in interpretability. We find that BoSsNet outperforms state-of-the-art models, with considerable improvements (> 10\%) on bAbI OOV test sets and other human-human datasets. We also systematically modify existing datasets to measure disentanglement and show BoSsNet to be robust to KB modifications.

    See publication

  • Beyond Anomaly Detection: Lookout for Pictorial Explanation

    ECML-PKDD 2018

    Why is a given point in a dataset marked as an outlier by an off-the-shelf detection algorithm? Which feature(s) explain it the best? What is the best way to convince a human analyst that the point is indeed an outlier? We provide succinct, interpretable, and simple pictorial explanations of outlying behavior in multi-dimensional real-valued datasets while respecting the limited attention of human analysts. Specifically, we propose to output a few pictures (focus-plots, ie., pairwise feature…

    Why is a given point in a dataset marked as an outlier by an off-the-shelf detection algorithm? Which feature(s) explain it the best? What is the best way to convince a human analyst that the point is indeed an outlier? We provide succinct, interpretable, and simple pictorial explanations of outlying behavior in multi-dimensional real-valued datasets while respecting the limited attention of human analysts. Specifically, we propose to output a few pictures (focus-plots, ie., pairwise feature plots) from a few, carefully chosen feature sub-spaces. The proposed LookOut makes four contributions: (a) problem formulation: we introduce an “analyst-centered” problem formulation for explaining outliers via focus-plots, (b) explanation algorithm: we propose a plot-selection objective and the LookOut algorithm to approximate it with optimality guarantees,(c) generality: our explanation algorithm is both domain- and detector-agnostic, and (d) scalability: LookOut scales linearly with the size of input outliers to explain and the explanation budget. Our experiments show that LookOut performs near-ideally in terms of maximizing explanation objective on several real datasets, while producing fast, visually interpretable and intuitive results in explaining groundtruth outliers from several real-world datasets.

    See publication

Courses

  • Analysis and Design of Algorithms

    COL351

  • Artificial Intelligence

    COL333

  • Computer Architecture

    COL216

  • Computer Networks

    COL334

  • Data Structure and Algorithms

    COL106

  • Deep Learning

    COL876

  • Design Practices

    COP290

  • Digital Image Analysis

    COL783

  • Digital Logic and Circuit Design

    COL215

  • Discrete Mathematical Structures

    COL202

  • Introduction to Computer Science

    COL100

  • Machine Learning

    COL774

  • Machine Learning

    COL761

  • Natural Language Processing

    COL772

  • Operating Systems

    COL331

  • Parallel and Distributed Programming

    COL380

  • Probability and Stochastic Processes

    MTL106

  • Programming Languages

    COL226

  • Theory of Computation

    COL352

Honors & Awards

  • BOSS Award

    IIT Delhi

    I received the BOSS award for the best Master's Project in Computer Science and Engineering in the graduating batch of 2019.

  • President's Gold Medal

    School Board

    I received the school's Gold Medal for finishing at the top of my class out of 300 students. I also received a scholarship of 50000 rupees for my academic achievements and excellence.

Test Scores

  • Bachelor's GPA

    Score: 9.2

    My cumulative GPA for my Bachelor's Degree

  • Master's GPA

    Score: 9.7

    My cumulative GPA for my Master's Degree

  • IIT-JEE

    Score: Rank 206

    Secured Rank 206 in the IIT-JEE and got a place in IIT Delhi

  • IIT-JEE Mains

    Score: Rank 43

    Got 336/360 marks in the IIT-JEE Mains examination of 2014

  • Class 12 Boards

    Score: 96.4 %

    I secured a total of 94% in my school cumulative of 11th and 12th and was the school topper. In my class 12th boards I got a total percentage of 96.4%

  • KVPY

    Score: AIR 8

    I secured All India Rank 8 in the prestigious KVPY examination

Languages

  • English

    Native or bilingual proficiency

  • Hindi

    Native or bilingual proficiency

  • Spanish

    Elementary proficiency

  • French

    Elementary proficiency

Recommendations received

1 person has recommended Nikhil

Join now to view

More activity by Nikhil

View Nikhil’s full profile

  • See who you know in common
  • Get introduced
  • Contact Nikhil directly
Join to view full profile

Other similar profiles

Explore more posts

Explore collaborative articles

We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.

Explore More

Others named Nikhil Gupta in United States

167 others named Nikhil Gupta in United States are on LinkedIn

See others named Nikhil Gupta

Add new skills with these courses

See all courses