NUWA RESEARCH RECORD
Reasoning models struggle to control their chains of thought, and that’s good
Original publication ↗ · BibTeX · RISTested models show low controllability; more reasoning and some RL training reduce it.
Research question
Can reasoning models deliberately shape the traces used for oversight?
Design & assumptions
CoT-Control pairs existing benchmark tasks with instructions constraining reasoning text; compares scale and reasoning budgets.
Findings
Tested models show low controllability; more reasoning and some RL training reduce it.
Limits & counterevidence
Instruction-following is a proxy for deliberate monitor evasion; low controllability alone cannot establish monitorability.
Source evidence
No structured numerical extraction is available in this reading package.
Reading scope & review
Selected original webpage sections; linked paper, appendix and data not fully audited.
Model-assisted reading; independent evidence review pending.
Check the original source ↗