MétaCan
Menu
Back to cohort
Record W2575334674

Improved heuristic and tie-breaking for optimally solving Sokoban

2016· article· en· W2575334674 on OpenAlexaff
André G. Pereira, Robert C. Holte, Jonathan Schaeffer, Luciana S. Buriol, Marcus Ritt

Bibliographic record

VenueInternational Joint Conference on Artificial Intelligence · 2016
Typearticle
Languageen
FieldComputer Science
TopicLogic, Reasoning, and Knowledge
Canadian institutionsUniversity of Alberta
Fundersnot available
KeywordsHeuristicsHeuristicComputer scienceMathematical optimizationConsistent heuristicDomain (mathematical analysis)MathematicsAlgorithmIncremental heuristic searchTheoretical computer scienceSearch algorithmBeam search
DOInot available

Abstract

fetched live from OpenAlex

We present a novel admissible pattern database heuristic (D) and tie-breaking rule (L) for Sokoban, allowing us to increase the number of optimally solved standard Sokoban instances from 20 to 28 and the number of proved optimal solutions from 25 to 32 compared to previous methods. The previously best heuristic for Sokoban (I) used the idea of an intermediate goal state to enable the effective use of pattern database heuristics in transportation domains, where the mapping of movable objects to goal locations is not fixed beforehand. We extend this idea to allow the use of multiple intermediate goal states and show that heuristic I is no longer effective. We solve this problem and show that our heuristic D is effective in this situation. Sokoban is a well-known single-agent search domain characterized by a large branching factor, long solution lengths, and the presence of unsolvable states. Given the exponential growth in the complexity of standard Sokoban instances, the increase in the number of optimally solved instances represents a major advance in our understanding of how to search in extremely large search spaces.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame distilled prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.

metaresearch head score (Codex)0.000
metaresearch head score (Gemma)0.001
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Theoretical or conceptual · Consensus signal: none
GenreCandidate signal: Methods · Consensus signal: none
Teacher disagreement score0.973
Threshold uncertainty score0.713

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.001
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.000
Bibliometrics0.0000.000
Science and technology studies0.0000.000
Scholarly communication0.0000.000
Open science0.0010.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0000.000

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.090
GPT teacher head0.309
Teacher spread0.219 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one teacher head, not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designTheoretical or conceptual
Domainnot available
GenreMethods

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations3
Published2016
Admission routes1
Has abstractyes

Explore more

Same venueInternational Joint Conference on Artificial IntelligenceSame topicLogic, Reasoning, and KnowledgeFrench-language works237,207