VQV Signal

ROBOTS & HARDWARE WATCH TECHNICAL

Five new robot models tested zero-shot on towel folding in RoboColosseum post

According to a post on X by @RoboColosseum, five robot models released after CoRL were given the same zero-shot towel-folding task on a bimanual YAM robot. Fable 5.2 and π0.5 showed a fold, Dream-1 left the towel bunched, MolmoAct2 held it with both grippers, and Gemini repositioned it without moving it. The post warns that latency issues mean timings are not directly comparable across models. It also offers academic labs a share of 10,000 free evaluation hours.

Source: Scobleizer News Pulse (X) · x.com Found via Robert Scoble’s Scobleizer News Pulse · a post on X, not independently verified Published 2026-11-19T20:36:00+00:00 Detected 2026-11-20T05:01:50+00:00
View original source

According to a post on X by @RoboColosseum, five robot models released after CoRL were given the same zero-shot towel-folding task on a bimanual YAM robot. Fable 5.2 and π0.5 showed a fold, Dream-1 left the towel bunched, MolmoAct2 held it with both grippers, and Gemini repositioned it without moving it. The post warns that latency issues mean timings are not directly comparable across models. It also offers academic labs a share of 10,000 free evaluation hours.

AI-assisted summary based on the listed source.

After CoRL's model releases, RoboColosseum gave five models the same zero-shot task on a bimanual YAM: Fable 5.2 and π0.5 showed a fold, Dream-1 left the towel bunched, MolmoAct2 held it with both grippers and Gemini repositioned without moving it. It warns latency issues make timings not directly comparable and...

Academic robotics labs are the group the post directly targets, since it offers them a share of 10,000 free evaluation hours. Anyone comparing these models should treat the results as one informal test, because the latency caveat limits direct timing comparisons. The snippet does not describe the full methodology, so the next thing to watch is whether more detailed results are published.

Hardware and robotics watchers may want to track whether this becomes a product, benchmark, or deployment signal.

Signal Strength 91% Technical label WATCH Public Interest 37 Category ROBOTS & HARDWARE Reader Depth TECHNICAL

Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.

Public Interest components
Recognizable Entity Score 0 Practical Impact Score 26 Novelty Interest Score 94 Consequence Score 18 Curiosity Score 52 Shareability Score 50

VQV surfaced this signal because it is recent, relevant to Open Source LLMs, connected to Scobleizer News Pulse (X).