Pulkit Madan

Papers

Total Citations

H-Index

About

Pulkit Madan is a researcher at the forefront of integrating large language models (LLMs) with creative and visual domains. Their primary research areas span natural language processing, computer vision, and generative AI, with a particular focus on extending the capabilities of autoregressive models beyond text. Madan’s most notable contribution is the pioneering work "Painter: Teaching Auto-regressive Language Models to Draw Sketches" (2023), which demonstrates how LLMs can be repurposed for image generation tasks by directly producing sketch-like outputs. This innovative approach bridges the gap between language understanding and visual creativity, opening new avenues for multimodal AI. While still early in its impact, with 2 citations to date, the work has garnered attention for its novel methodology and potential applications in creative tools, education, and human-computer interaction. Madan’s research highlights the versatility of LLMs, showing that models trained on text can learn to generate structured visual content without specialized architectures. By pushing the boundaries of what language models can do, Madan is contributing to a future where AI systems seamlessly blend linguistic and visual reasoning, inspiring further exploration into cross-modal generative tasks.