Weights or Skills? A Survey of Robot-Learning Techniques: from Action-Predicting Weights to Robots that Write their Own Skills
2026-08-03 • Robotics
RoboticsArtificial Intelligence
AI summaryⓘ
The authors explain that robot learning mainly takes two approaches: one where robots have fixed 'brains' that handle tasks, and another where robots create and improve their own skills as code. They organize methods that use code by how much the robot can improve itself, highlighting that very few systems fully combine all self-improvement techniques. They also clarify that the term 'skill' means different things in this field and only code-based skills can self-improve without retraining. Finally, the authors discuss how robot skill marketplaces currently only share fixed skills, raising challenges like adapting to new robots and ensuring safety.
robot learningpolicyvision-language-action (VLA) modelscode-as-policyself-improvementskill memoryprogram synthesisreinforcement learningskill economyrobot skill marketplaces
Authors
Gaytri Jena, Kapil Wanaskar, Vinija Jain, Aman Chadha, Vasu Sharma, Amitava Das
Abstract
Robot learning is splitting into two bets: policies that bake competence into frozen weights (vision-language-action, or VLA, models), and agents that write and refine their own executable skills as code. This survey organises the field around that axis of weights versus skills. Its central analytical contribution is a deep-dive that arranges code-as-policy methods by their degree of self-improvement, from zero-shot program synthesis, through closed-loop self-repair and persistent skill memory, to the sparsely populated cell in which execution feedback, skill memory, and evolutionary search combine into one open-ended loop; only a few very recent systems (for example ASPIRE, ENPIRE, and RoboClaw) occupy that cell. We map the complementary "skills" pole, from unsupervised reinforcement-learning skill discovery to large-language-model skill libraries, and show that the word "skill" is used in at least five distinct senses, of which only the code sense self-improves without gradient updates. We then connect the taxonomy to the emerging skill economy: commercial robot-skill marketplaces now distribute one-tap skills across robots but ship only static playback, which surfaces open problems of adaptation, cross-embodiment portability, provenance, safety verification, composition, and standardisation. This is a deliberately focused survey. Rather than cataloguing the field exhaustively, it examines 77 representative systems across six technique families through one taxonomy and a set of contrast tables, and it supplies operational definitions of the self-improvement mechanisms together with a statement of what each family cannot do.