Projects
Projects
Viveka
Viveka
With the aim of reducing hallucinations in Large Language Models (LLMs), this project seeks to address this persistent challenge through the emerging field of mechanistic interpretability - a discipline that delves into the inner workings of neural networks to understand why models produce the outputs they do. By dissecting and analyzing internal representations and activation patterns, the project aspires to uncover the mechanisms behind factual inconsistencies and build methods that make LLMs more reliable and grounded. The project has been awarded grants from VCs and ExceptionRaised!
With the aim of reducing hallucinations in Large Language Models (LLMs), this project seeks to address this persistent challenge through the emerging field of mechanistic interpretability - a discipline that delves into the inner workings of neural networks to understand why models produce the outputs they do. By dissecting and analyzing internal representations and activation patterns, the project aspires to uncover the mechanisms behind factual inconsistencies and build methods that make LLMs more reliable and grounded. The project has been awarded grants from VCs and ExceptionRaised!
Deep Recall
Deep Recall
Deep Recall is a collaborative project between the AI Club and the Archive Center of IIT Madras, proposed by the Director of IITM. It aims to build a scalable system that indexes and retrieves historical records through embeddings and semantic search, while ensuring data privacy. The project integrates OCR, knowledge graphs, and advanced retrieval frameworks like LangChain, GraphRAG, and FAISS, with the long-term goal of evolving into a deployable startup solution for institutions worldwide.
Deep Recall is a collaborative project between the AI Club and the Archive Center of IIT Madras, proposed by the Director of IITM. It aims to build a scalable system that indexes and retrieves historical records through embeddings and semantic search, while ensuring data privacy. The project integrates OCR, knowledge graphs, and advanced retrieval frameworks like LangChain, GraphRAG, and FAISS, with the long-term goal of evolving into a deployable startup solution for institutions worldwide.
Speechseek
Speechseek
Project SpeechSeek aims to develop an advanced speaker diarization system with integrated emotion and language detection capabilities, designed to handle complex code-mixed speech and support multiple Indic languages, addressing a highly intriguing and challenging problem in the rapidly evolving world of neural audio processing and speech understanding.
Project SpeechSeek aims to develop an advanced speaker diarization system with integrated emotion and language detection capabilities, designed to handle complex code-mixed speech and support multiple Indic languages, addressing a highly intriguing and challenging problem in the rapidly evolving world of neural audio processing and speech understanding.