Teahose.
SIGN IN
NEW HERE — WHAT TEAHOSE DOES
We read the entire AI & tech firehose — so you don't have to.
PODPodcastsAll-In, No Priors, Acquired…
NEWNewslettersStratechery, Newcomer…
PAPPapersPhysical AI research
PHProduct Huntdaily launches
VCInvestor ScoutSequoia, a16z, Benchmark…
CLAUDE DISTILLS →
7 reads, 30 sec each — free, 6 AM ET.
+ a live graph of the companies, people & themes underneath.
HOME/PEOPLE/WILBERT PUMACAY
// PERSON

Wilbert Pumacay

ROLE RESEARCHERMENTIONS 1LAST SEEN JUNE 1, 2026
// BIO

Wilbert Pumacay is a computer science researcher specializing in reinforcement learning for robotics. He is a graduate student at Universidad Católica San Pablo in Arequipa, Peru, where he works on robotic manipulation. He is best known as a co-author of The Colosseum, a simulation benchmark for evaluating generalization in robotic manipulation that introduces a systematic perturbation taxonomy including manipulation object and receiver object categories.

// RECENT MENTIONS
// SIGNALS
1 SIGNAL
01
mention·arXiv Physical AI·JUNE 1, 2026

Wilbert Pumacay — Co-author of the original Colosseum benchmark (reference [32]), which Colosseum V2 extends. His prior work established the perturbation taxonomy and terminology (MO/RO) adopted in V2.

Source

AI-extracted from podcast / newsletter / paper summaries. May contain errors.