Model techniques map
Techniquesevaluation

specific method · filed under evaluation

avg@k

Mean pass@1 averaged over multiple attempts per task, with the number of attempts specified by k.

Also called average over 3 rollouts (avg@3), avg@3.

sources
2
models
2
labs adopt it
2
strongest
used

How sources treat it

One count per evidence span, weakest treatment to strongest.

used 2

Documented in

Evidence

2 spans quoted from the sources, strongest treatment first.

reporting avg@3 over three rollouts per task

usedunclearin GLM-5.3Z.ai

We report mean pass@1 averaged over multiple attempts per task (avg@k), with k per benchmark listed below.

usedevaluation onlyPoolside

Filed alongside

Other methods under evaluation.