We consider the problem of learning propositional STRIPS world models from action traces alone, using a deep learning architecture (transformers) and gradient descent. The task is cast as a supervised next token prediction problem where the tokens are the actions, and an action $a$ may follow an action sequence if the hidden effects of the previous actions do not make an action precondition of $a$ false. We show that a suitable transformer architecture can faithfully represent propositional STRIPS world models, and that the models can be learned from sets of random valid (positive) and invalid (negative) action sequences alone. A number of experiments are reported.

Consider the problem of learning a lifted STRIPS model of the sliding-tile puzzle from random state-action traces where the states represent the location of the tiles only, and the actions are the labels up, down, left, and right, with no arguments. Two challenges are involved in this problem. First, the states are not full STRIPS states, as some predicates are missing, like the atoms representing the position of the ``blank''. Second, the actions are not full STRIPS either, as they do not reveal all the objects involved in the actions effects and preconditions. Previous approaches have addressed different versions of this model learning problem, but most assume that actions in the traces are full STRIPS actions or that the domain predicates are all observable. The new setting considered in this work is more ``realistic'', as the atoms observed convey the state of the world but not full STRIPS states, and the actions reveal the arguments needed for selecting the action but not the ones needed for modeling it in STRIPS. For formulating and addressing the learning problem, we introduce a variant of STRIPS, which we call STRIPS+, where certain STRIPS action arguments can be left implicit in preconditions which can also involve a limited form of existential quantification. The learning problem becomes the problem of learning STRIPS+ models from STRIPS+ state-action traces. For this, the proposed learning algorithm, called SYNTH, constructs a stratified sequence (conjunction) of precondition expressions or ``queries'' for each action, that denote unique objects in the state and ground the implicit action arguments in STRIPS+. The correctness and completeness of SYNTH is established, and its scalability is tested on state-action traces obtained from STRIPS+ models derived from existing STRIPS domains.

A popular approach for sequential decision-making is to perform simulator-based search guided with Machine Learning (ML) methods like policy learning. On the other hand, model-relaxation heuristics can guide the search effectively if a full declarative model is available. In this work, we consider how a practitioner can improve ML-based black-box planning on settings where a complete symbolic model is not available. We show that specifying an incomplete STRIPS model that describes only part of the problem enables the use of relaxation heuristics. Our findings on several planning domains suggest that this is an effective way to improve ML-based black-box planning beyond collecting more data or tuning ML architectures.