top of page

Hi, I hope you are having a Good Day!


I am a lazy researcher who works towards making NLP systems more robust and have majorly contributed in the following areas of NLP and ML:


Safety of LLM with a focus on robust alignment evaluations and algorithms. Major contributions include Red-Eval and Unalignment.

Analysis of knowledge encoded in NLP models that could be helpful for (let's say) gender debiasing a system and safety issues in LLMs.

Designing theoretically grounded algorithms to make models more robust and explainable.

Google Scholar       GitHub       CV


Official  | Personal

Work Experience
After gaining invaluable learning experience at BITS Pilani, I have amassed research ethics and attitude from various groups such as CEERI, University of Alberta, and National University of Singapore. When it comes to most valuable research experiences in the list of invaluable ones (kind of comparing two infinitely large numbers), I can not think beyond my research tenure at Salesforce research Asia and Amazon Science, California. On a side note, do look at the Mathology section for my side interests.

Most of my efficient learning in the domain of AI came after I joined SUTD as a Ph.D. candidate in 2020. One year of the worthy ride on the Ph.D. tuk-tuk carried me from the applied stop of the machine learning to the stop where I found interest in theoretical considerations of NLP models, i.e., mathematical analysis of learned representations and how to perform efficient distillation of the encoded knowledge. During the ride, I have shared the space with a few people (co-authors) without whom the journey would not be this exciting. Optimistically and without being overambitious (I had to check spellings of both the confusing words), I am generously waiting for the next few stops before the tuk-tuk says goodbye to me.
LLM Safety (ACL 2024)

Language Models are Homer Simpson! Safety Re-Alignment of Fine-tuned Language Models through Task Arithmetic

(Rishabh Bhardwaj, Do Duc Anh, Soujanya Poria)

Unalignment in LLMs

Language Model Unalignment: Parametric Red-teaming to Expose Hidden Harms And Biases

(Rishabh Bhardwaj, Soujanya Poria)

EMNLP (2023)

Adapter Pruning using Tropical Characterization

(Rishabh Bhardwaj, Tushar Vaidya, Soujanya Poria)

EMNLP (2023)

kNN-CM: A Non-parametric Inference-Phase Adaptation of Parametric Text Classifiers

(Rishabh Bhardwaj*, Yingting Li*, Navonil Majumdar, Soujanya Poria)

ACL (2023)

Adaptation Approaches for Nearest Neighbor Language Models

(Rishabh Bhardwaj, George Polovets, Monica Sunkara)

ACL (2023)

ReMask: A Robust Information-Masking Approach for Domain Counterfactual Generation

(Pengfei Hong*, Rishabh Bhardwaj*, Navonil Majumdar, Somak Aditya, Soujanya Poria)

EMNLP (2022)

Vector-Quantized Input-Contextualized Soft Prompts for Natural Language Understanding

(Rishabh Bhardwaj* and Amrita Saha* and Steven C.H. Hoi and Soujanya Poria)

COLING (2022)

KNOT: Knowledge Distillation using Optimal Transport for Solving NLP Tasks

(Rishabh Bhardwaj and Tushar Vaidya and Soujanya Poria)

ACL (2021)

More Identifiable yet Equally Performant Transformers for Text Classification

(Rishabh Bhardwaj and Navonil Majumder and Soujanya Poria and Eduard Hovy)

Cognitive Computation (2021)

Investigating Gender Bias in BERT

(Rishabh Bhardwaj and Navonil Majumder and Soujanya Poria)

Neural Computing and Applications (2020)

Improving Aspect-Level Sentiment Analysis with Aspect Extraction

(Navonil Majumder* and Rishabh Bhardwaj* and Soujanya Poria and Amir Zadeh and Alexander Gelbukh and Amir Hussain and Louis-Philippe Morency)

ACL (2019)

Twitter Homophily: Network-Based Prediction of User’s Occupation

(Jiaqi Pan* and Rishabh Bhardwaj* and Wei Lu and Hai Leong Chieu and Xinghao Pan and Ni Yi Puay)

In total, I have been fortunate to get a chance in co-authoring more than 20 research papers
related to applied machine learning or designing algorithms to make it robust.

I have been a part of reviewing committee at top-tier NLP conferences such as AAAI 2023, EMNLP 2023, ACL 2023, AAAI 2022, ACL 2022, EMNLP 2022, EMNLP 2021, ACL 2021, NAACL 2021, ACL 2020, and Journals such as Information Fusion and Neural Computing and Applications.


Sentiment Analysis - 2nd rank
Shopee Sentiment Analysis Challenge.

Product Detection - 2nd rank

Shopee Product Detection Challenge.


Thanks to my mentor Soujanya for pushing my boundaries. I have experience in programming languages such as Java and C++. Most of my current experiments utilize Python and PyTorch. I use Tensorflow for industry-related projects.

Teaching: I have given a 2-hour online lesson covering the need for PyTorch over NumPy in-depth. I am attaching the video if you find it useful!

PyTorch Basics: A student's perspective.

PyTorch Basics: A student's perspective.

Play Video

Looking forward to answering your email.

Thanks for submitting!

I am always grateful to my brother, parents, friends, mentors, and Declare lab to provide me all the support!
bottom of page