Model techniques map
Techniquesevaluation

specific method · filed under evaluation

Refusal behavior quantification

Separating refusal-based failures from other failure modes and reporting refusal rates.

Also called Quantified refusal behavior.

source
1
model
1
lab adopt it
1
strongest
used

How sources treat it

One count per evidence span, weakest treatment to strongest.

used 1

Documented in

Evidence

1 span quoted from the sources, strongest treatment first.

Explicitly separated refusal-based failures from other failure modes and reported pre- and post-naughtification rates.

usedevaluation onlyin gpt-oss-120b and gpt-oss-20bOpenAI

Filed alongside

Other methods under evaluation.