AI brief
A research update reports results from a replication of Serrano et al., 2026, finding that models Astra and Sol can take low-probability actions without chain-of-thought reasoning.
Why it matters: The author says the results are released in the hope of inspiring similar or follow-up experiments.
Written by AI from LessWrong's published text.
Source
“This is a research update from a replication of Serrano et al., 2026 . I’m releasing these results now because I hope it inspires others to run similar / followup experiments.”
Interpretation or community commentary rather than straight reporting.
Read original story ↗