Safety
stunning confirmation of my central work
stunning confirmation of my central work Transformers struggle to generalize to tasks they were not explicitly trained on. Instead, we propose in 2026 that it is the job of the harness to generalize t
stunning confirmation of my central work Transformers struggle to generalize to tasks they were not explicitly trained on. Instead, we propose in 2026 that it is the job of the harness to generalize through composition. We observe a powerful property when training RLMs: for tasks with shared structure that look differen…
Related
- @GaryMarcus the whole AI bubble depends on investors not understanding that the progress on coding harnesses DOES NOT GENERALIZE to other le…
- neurosymbolic world models – exactly what I argued for in 2020 in the Next Decade in AI – for the win.
Source: Gary Marcus (X) | 2026-09-27