#skill-evolution

0 stories this week · 1 this month · 3 in 6 months · mostly AI

Stories per day · 30 days

3 stories, newest first · one card per story

EvoGuard uses pairwise rejection sampling on private test sets to select non-degrading skills at low latency.

EVISKILL uses replayable evidence so LLM agents can refine skills without updating model parameters.

Rep2Skill revises LLM agent skills from internal representation trajectories and beats text-only self-evolution.

Stories tagged #skill-evolution