Devansh Agarwal

Guest Author

Devansh Agarwal currently works as an ML engineer at FAANG, where he develops cutting-edge AI solutions for enterprise modernization. His journey in technology spans from building India’s first VR product for the architecture industry to leading machine learning initiatives at Amazon. In 2023, Devansh was selected as one of six Ignite Fellows by Cornell University, receiving a $240,000 award to transform his research into a commercial venture. This prestigious fellowship recognized his innovative work in AI and human-computer interaction. His work has been featured in major media outlets including Forbes, Hackster.io, and TechRadar.

RLVR DDM

Why reinforcement learning plateaus without representation depth (and other key takeaways from NeurIPS 2025)

Every year, NeurIPS produces hundreds of impressive papers, and a handful that subtly reset how practitioners think about scaling, evaluation and system design. In 2025, the most consequential works weren't about a single breakthrough model. Instead, they challenged fundamental assumptions that academicians and corporations have quietly relied on: Bigger models mean better reasoning, RL creates new capabilities, attention is “solved” and generative models inevitably memorize.

Maitreyi Chatterjee,Devansh Agarwal