MétaCan
Menu
Back to cohort
Record W6930627837 · doi:10.5281/zenodo.15586709

Scenes partitioning and annotations of Super Mario Bros. levels.

2025· other· en· W6930627837 on OpenAlexaff

Bibliographic record

VenueZenodo (CERN European Organization for Nuclear Research) · 2025
Typeother
Languageen
FieldMedicine
TopicFungal Biology and Applications
Canadian institutionsUniversité de MontréalInstitut Universitaire de Gériatrie de Montréal
Fundersnot available
KeywordsPoint (geometry)Position (finance)Integer (computer science)Integer programmingPerspective (graphical)

Abstract

fetched live from OpenAlex

Resources used to split Super Mario Bros. levels into succcessive "scenes" Each level map was first obtained from NesMaps and added to the mario_scenes_manual_annotations.pdf file. Using the stable-retro GUI, we obtained the X positions corresponding to the start and end of each scene. Then for each scenes, we identified game design patterns as described in Dahlskog & Togelius, 2012. We aggregated all these informations in the scenes_mastersheet.tsv file. Note : Underwater and Castle levels were ignored in our analysis because they have a slightly different gameplay than the regular level, and make use of different game design patterns. Bonus zones and water sections are annotated as such, but their patterns weren't identified. The scenes mastersheet This TSV file contains informations related to all the scenes identified in Super Mario Bros. levels. It contains one row per scene, 3 columns to identify the scene, an Entry and Exit point columns, and one columns per pattern. These columns contain the following information : - World : The world ID, an integer between 1 and 8.- Level : The level ID, an integer between 1 and 3.- Scene : The scene ID, an integer.- Entry point : The X position corresponding to the beginning of the scene. An integer.- Exit point : The X position corresponding to the ending of the scene. An integer. Design patterns (from Dahlskog & Togelius 2012). The values can be 0 (absence of the corresponding pattern) or 1 (presence of the corresponding pattern) : - Enemy : A single enemy- 2-Horde : Two enemies together- 3-Horde : Three enemies together- 4-Horde : Four enemies together- Roof : Enemies underneath a hanging platform making Mario bounce in the ceiling- Gap : Single gap in the ground/platform- Multiple gaps : More than one gap with fixed platforms in between- Variable gaps : Gap and platform width is variable- Gap enemy : Enemies in the air above gaps- Pillar gap : Pillar (pipes or blocks) are placed on platforms between gaps- Valley : A valley created by using vertically stacked blocks or pipes but without Piranha plant(s)- Pipe valley : A valley with pipes and Piranha plant(s)- Empty valley : A valley without enemies- Enemy valley : A valley with enemies- Roof valley : A valley with enemies and a roof making Mario bounce in the ceiling- 2-Path : A hanging platform allowing Mario to choose different paths- 3-Path : 2 hanging platforms allowing Mario to choose different paths- Risk/Reward : A multiple path where one path have a reward and a gap or enemy making it risky to go for the reward- Stair up : A stair going up- Stair down : A stair going down- Empty stair valley : A valley between a stair up and a stair down without enemies- Enemy stair valley : A valley between a stair up and a stair down with enemies- Gap stair valley : A valley between a stair up and a stair down with gap in the middle We added several patterns in order to annotate key sections of the level : - Reward : Rewards without immediate danger- Moving platform : Platform moving vertically or horizontally- Flagpole : End of the level- Beginning : Beginning of the level- Bonus zone : Hidden zone without enemies- Waterworld : A special hidden zone with Waterworld gameplay

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)0.000
Version: codex-gemma-dda1882f352aValidation status: machine_predicted_unvalidated
Candidate categoriesInsufficient payload (model declined to judge)
Consensus categoriesInsufficient payload (model declined to judge)
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Not applicable · Consensus signal: Not applicable
GenreCandidate signal: Other · Consensus signal: Other
Teacher disagreement score0.169
Threshold uncertainty score1.000

Codex and Gemma teacher scores by category

CategoryCodexGemma
Metaresearch0.0000.000
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 science0.0000.000
Research integrity0.0000.000
Insufficient payload (model declined to judge)0.0210.001

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 head0.036
GPT teacher head0.275
Teacher spread0.238 · 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; both teacher heads agree on what is shown here.

Study designNot applicable
Domainnot available
GenreOther

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

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same venueZenodo (CERN European Organization for Nuclear Research)Same topicFungal Biology and ApplicationsFrench-language works237,207