Sandeep Srinivasa

Early thoughts of java architecture

1

# Customized Observable base class let's create an abstraction `EdgeChain` on top of RxJava3, which will have customized versions of map, zip, and subscribe functions. We'll create two subclasses of...

implement watermarking

1

https://arxiv.org/abs/2301.10226

we present a novel perspective for solving knowledge-intensive tasks by replacing document retrievers with large language model generators. We call our method generate-then-read (GENREAD), which first prompts a large language...

high costs for most researchers and companies seeking to fine-tune these models for their specific use cases or domains. Moreover, users who can afford to fine-tune must provide their private...

automatically index postgresql to vector db

2

instead of me pushing data to be indexed, it connects to a postgresql db, inserts a hook where any update to my postgres is automatically processed

synthetic health data

2

https://synthetichealth.github.io/synthea/

https://arxiv.org/pdf/2212.08073.pdf

implement hyde

2

https://twitter.com/darrenangle/status/1652014961806196745 https://wfhbrian.com/revolutionizing-search-how-hypothetical-document-embeddings-hyde-can-save-time-and-increase-productivity/ https://arxiv.org/pdf/2212.10496.pdf > > Semantic search on embeddings is hard to get right. Embedding long documents is a challenge. User queries are a challenge, if a user provides an...

https://arxiv.org/pdf/2210.09150.pdf

https://github.com/debjitpaul/refiner https://arxiv.org/abs/2304.01904 LLMs Getting Fine-Grained Reasoning Feedback Improves Performance -Critic model trained to provide feedback reasoning LLM uses to iteratively improve intermediate arguments -Easily beats baselines -Even w/ GPT3.5, critic...