Loading…
The reported tests suggest weight-based behavior identification can work beyond the paper’s highlighted model, but the evidence here is a small, unverified set of examples.
Limits of the evidence: This is a single author's account of a small number of tests; the source provides no independent replication or systematic accuracy evaluation.
What to watch
Assessment revised Nov 4, 2026
Qwen 3.8 27B on my laptop can interpret model weights, too
Simon WillisonNov 3, 2026First report
First assessment, from Simon Willison
No maturity ladder applies to this desk.
Sens.ai aggregates and assesses published reporting. The assessment above is machine generated; the original sources are authoritative.