👋 Hey, I'm Mridul Sharma
Exploring large language models to enhance generalization and understand them.
I am Machine Learning Engineer at Keeper Dating Inc. where I currently work on developing retrieval + ranking algorithms for matchmaking and also on large language model distillation.
My interests include:
- •Large Language Models (LLMs)
- •Reinforcement Learning
- •Parallel Processing
- •AI for Social Good
Featured Publications
Selected works from my research
Confirmation bias: A challenge for scalable oversight
We conducted two studies examining the performance of simple oversight protocols where evaluators know that the model is correct most of the time, but not all of the time.
Development of Pre-Trained Transformer-based Models for the Nepali Language
We collected 27.5 GB of Nepali text data, approximately 2.4x larger than any previously available Nepali language corpus. Leveraging this data, we pre-trained three different models i.e., BERT, RoBERTa, and GPT-2, exclusively for the Nepali Language.
Consolidating and Developing Benchmarking Datasets for the Nepali Natural Language Understanding Tasks
We introduce eight new datasets, creating a new benchmark, the Nepali Language Understanding Evaluation (NLUE) benchmark, which covers a total of 12 tasks for evaluating the performance of models across a diverse set of Natural Language Understanding (NLU) tasks. The added tasks include single-sentence classification, similarity and paraphrase tasks, and Natural Language Inference (NLI) tasks. On evaluating the models using added tasks, we observe that the existing models fall short in handling complex NLU tasks effectively.
Recent Blog Posts
Thoughts and insights on research and technology
