Aditya Kusupati

@adityakusupati

Been places..... Done things.... Wandering @GoogleDeepMind Enjoying little things w/ @_keysarasara

following
followers

Claims that Argon skips reasoning or leaks it as plain text

A user says Argon often doesn't reason at all or leaks all of its reasoning as plain text.

22 Sources, 5d ago, updated 5d ago

‘Tokenizer Myths and Where To Find Them’ talk seeks early feedback

The speaker invites people to DM for slides and says the work is unrelated to their startup.

12 Sources, 18d ago, updated 18d ago

Gemini reinforcement-learning work reportedly brings bug fixes and improvements

Arthur Douillard describes work since March to scale reinforcement learning for Gemini, with bug fixes, improvements and more still to build.

28 Sources, 19d ago, updated 19d ago

Google introduces Gemini 4 Argon with limited cyber rollout

Google announces a 1 million-token output limit, up from 64,000, for its new model. Initial access goes to trusted cyber defenders through the Fairwind Program.

333 Sources, 15d ago, updated 15d ago

Jeff Dean and Colleagues Launch Discovery Loop

Four former Google researchers found a public benefit corporation to automate machine learning and science.

Jeff Dean and Colleagues Found Discovery Loop

221 Sources, 76d ago, first seen 76d ago

Featured in

Claims that Argon skips reasoning or leaks it as plain text

‘Tokenizer Myths and Where To Find Them’ talk seeks early feedback

Gemini reinforcement-learning work reportedly brings bug fixes and improvements

Google introduces Gemini 4 Argon with limited cyber rollout

Jeff Dean and Colleagues Launch Discovery Loop

Featured in

Claims that Argon skips reasoning or leaks it as plain text

‘Tokenizer Myths and Where To Find Them’ talk seeks early feedback

Gemini reinforcement-learning work reportedly brings bug fixes and improvements

Google introduces Gemini 4 Argon with limited cyber rollout

Jeff Dean and Colleagues Launch Discovery Loop