Rahul Chand

RahulRahul is the name Buddha gave to his son (Also means Silver) ChandChand means "moon" in Sanskrit

राहुल चंद

MS CS · Stanford University

Hi, I am Rahul. I recently completed my MS CS from Stanford 🌲 . Previously I was an intern at the GEAR Lab at NVIDIA Research training GR00T N1.6 and working on offline RL. . Before that, I was a summer intern at NVIDIA's post-training Nemotron LLM team working on looped transformer, and contributed to Nemo-RL (async RL).

I was also part of the ILIAD group (Prof. Dorsa Sadigh's lab). Prior to Stanford 🌲 I was a Research Fellow at Microsoft Research on Transformer Compression & Extreme Classification. My current interests are in RL+LLM and Robotics.

At Stanford 🌲 I have been a TA for:

In the futurefuture meme I look forward to getting replaced by AGI and work as a Wallace design protein farmerwallace design.

In my previous life (most of which I have forgottenforgotten memory), I completed my undergrad in CS from BITS Pilani and wanted to win India its first Oscar Award for Best International Film. For more details, check my CV or my favourite movies of all time.

Publications

Data Retrieval with Importance Weights for Few-Shot Imitation Learning

Amber Xie, Rahul Chand, Dorsa Sadigh, Joey Hejna

CoRL 2025 Oral · Conference on Robot Learning

Enhancing Tail Performance in Extreme Classifiers by Label Variance Reduction

Anirudh Buvanesh*, Rahul Chand*, Yashoteja Prabhu, Manish Gupta, Manik Verma (* = Equal Contribution)

ICLR 2024 · International Conference on Learning Representations

CapsFlow: Optical Flow Estimation with Capsule Networks

Rahul Chand, Rajat Arora, Ram Prabhakar Venkatesh Babu

Open Source

Isaac-GR00T 6000+ ⭐ · Contributor

NVIDIA's open platform for humanoid robot learning — training, fine-tuning, and deploying GR00T foundation models.

gpu_poor 1400+ ⭐

Tool to check vRAM & token/s requirement for any LLM for consumer hardware. Supports ggml, HuggingFace, bitsandbytes, QLoRA & gradient checkpointing. Used over 120k+ times by 20k+ users.

Starter tutorial for inference with LLaMA in C.

Nemo-RL 1k+ ⭐ · Contributor

NeMo RL is NVIDIA's open-source post-training library.

Nemo-Gym 70+ ⭐ · Contributor

NeMo Gym is a framework for building reinforcement learning environments to train LLMs.

Stanford Stanford
NVIDIA NVIDIA
Microsoft Research Microsoft Research
IISc IISc
IIRS IIRS
BITS Pilani BITS Pilani