AI Research & Publications
Seminal neural network papers, open-source Indic benchmark methodologies, reinforcement learning discoveries, and government AI ethics whitepapers.
DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
Pioneered pure reinforcement learning without supervised fine-tuning (DeepSeek-R1-Zero), proving that large language models naturally develop self-verification, chain-of-thought exploration, and AHA-moments when incentivized with outcome-based rewards.
IndicTrans2: Towards High-Quality and Accessible Machine Translation for all 22 Scheduled Indian Languages
A monumental open-source machine translation architecture covering all 22 scheduled Indian languages across 44 combinations, drastically closing the low-resource benchmark gap for Indic digital sovereignty.
Constitutional AI: Harmlessness from AI Feedback
Introduced Reinforcement Learning from AI Feedback (RLAIF) guided by a written constitution of human values, eliminating the requirement for human labelers to view traumatic or toxic outputs during safety alignment.
Attention Is All You Need: The Transformer Architecture
The foundational scientific paper that introduced the self-attention mechanism, replacing recurrent and convolutional neural networks and giving birth to modern Large Language Models and Generative AI.