SkillRL

Prior framework recursively expanding a skill library through reinforcement learning

Neighborhood — ranked by edge-count

paper

thinker

Peng Xia
studies
Author of SkillRL, a skill-level self-evolution framework cited as prior work

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

Skill-Load Rateconcept0.711
The fraction of trajectories in which an agent actively loads at least one skill into its context
ReflCtrlframework0.699
The proposed framework for probing and steering self-reflection behavior in reasoning LLMs via representation engineering
Reinforcement learning (RL)concept0.697
Machine learning paradigm where agents learn to maximize cumulative reward through interaction.
A skill works at multiple reinforcing scales: individual command, skill invocation, and user project.claim0.688
Skill-Load Rate Measurementmethod0.685
Named metric measuring the fraction of trajectories in which a model actively loads at least one skill into its context
steppermethod0.684
Primitive signal transformer implementing sample-and-hold; transforms event source to continuous piecewise-constant signal.
ReActframework0.682
Prior framework for synergizing reasoning and acting in LLM agents, foundational to agent harness concept
Learningconcept0.682
Inference of parameters encoding contingencies of the world (e.g., likelihood matrix A) at slower timescale than perception.