Biswadip Dey

Senior Key Expert @ Siemens Technology | Research Focus: Physics-aware AI | he/him

[Biswadip Dey @DBiswadip]

@ylecun Wow! I didn’t know that you and AT&T Labs were behind djvu! 🙏🏼

[Biswadip Dey @DBiswadip]

@AndrewYNg 1) “Dissociating language and thought in large language models: a cognitive perspective” https://t.co/fwchVLyZ9x
2) “On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?” https://t.co/uzUxji6znV

[Biswadip Dey @DBiswadip]

@amitra_90 @MountainOfMoon Nope, this is from a neighborhood in Nimta/Birati… I saw it while going home from the airport…

[Biswadip Dey @DBiswadip]

If you are at #NeurIPS2022 , please stop by our poster at the ML4PS Workshop today. In this work, we have used a PINN for modeling deformation in elastic-viscoplastic solid materials.

[Biswadip Dey @DBiswadip]

@B1ar2n3a @MountainOfMoon I am going through the same process.😢 I have a conference travel to Berlin, but cannot go because I could not get a timely appointment for my Schengen visa. The wait times are beyond ridiculous.

[Richard Sutton @RichardSSutton]

If you take all the fields that study intelligent decision making—from neuroscience to AI, psychology to control theory, economics to operations research—do their theories have much in common? I think so, as I explain in this new short paper: https://t.co/F1HgNDp5xo

[Biswadip Dey @DBiswadip]

If you are attending #neurips2021, please check out our work on "Extending Lagrangian and Hamiltonian Neural Networks with Differentiable Contact Models" in Poster Session 6 (Today | 11:30 am - 1:00 pm EST). We will be at Spot E1 in Virtual World.
Looking forward to seeing you!

[Biswadip Dey @DBiswadip]

Excited to share our work that extends the scope of Hamiltonian/Lagrangian based Energy-conserving Neural Networks with differentiable contact models. We will present it at #NeurIPS2021. Paper:https://t.co/6tBkTkHNlA Code:https://t.co/7RltUusMb4 Video:https://t.co/BtiYSHkZ20
1/🧵

[Biswadip Dey @DBiswadip]

@HyunSooPark_UMN @SenguptRoni this might be of interest to you

[Biswadip Dey @DBiswadip]

Finally Kudos to amazing collaborators Desmond Zhong and Amit Chakraborty!
Please check-out the paper to learn more about the interpretability, sample efficiency, scalability, and robustness aspect of the proposed approach.
-- end of🧵

[Biswadip Dey @DBiswadip]

Excited to share our work that extends the scope of Hamiltonian/Lagrangian based Energy-conserving Neural Networks with differentiable contact models. We will present it at #NeurIPS2021. Paper:https://t.co/6tBkTkHNlA Code:https://t.co/7RltUusMb4 Video:https://t.co/BtiYSHkZ20
1/🧵

[Biswadip Dey @DBiswadip]

Inspired by DiffTaichi (Hu et al. ICLR 2020), we also demonstrate the framework on the Billiard task. Here, the goal is to find the initial position and velocity of the white ball such that the blue ball hits the black target at a specified time.
14/🧵