pilipolio/chess-positions-extractor

Model Card for chess-concept-probe

Model Details

Model Description

Trained multi-label classifier for detecting 12 chess concepts (discovered_attack, exposed_king, fork, initiative, mating_threat, outpost, passed_pawn, pin, sacrifice, skewer, weak_square, zugzwang) from LC0 layer activations (block3/conv2/relu).

Trained from pilipolio/chess-sandbox@df742cf.

• Developed by: chess-sandbox
• Funded by [optional]: [More Information Needed]
• Shared by [optional]: [More Information Needed]
• Model type: Multi-label tabular classifier (scikit-learn LogisticRegression)
• Language(s) (NLP): en
• License: mit
• Finetuned from model [optional]: lczerolens/maia-1500

Model Sources [optional]

• Repository: https://github.com/pilipolio/chess-sandbox
• Paper [optional]: [More Information Needed]
• Demo [optional]: [More Information Needed]

Uses

Direct Use

Extract chess concepts from positions. See https://github.com/pilipolio/chess-sandbox for usage examples.

Downstream Use [optional]

[More Information Needed]

Out-of-Scope Use

Not suitable for non-chess domains or positions outside training distribution.

Bias, Risks, and Limitations

[More Information Needed]

Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.

How to Get Started with the Model

Use the code below to get started with the model.

See https://github.com/pilipolio/chess-sandbox README for complete examples.

Training Details

Training Data

Dataset: pilipolio/chess-positions-concepts

Training Procedure

Preprocessing [optional]

[More Information Needed]

Training Hyperparameters

• Training regime: [More Information Needed]

Speeds, Sizes, Times [optional]

[More Information Needed]

Evaluation

Testing Data, Factors & Metrics

Testing Data

Same as training data. See pilipolio/chess-positions-concepts

Factors

[More Information Needed]

Metrics

Multi-label metrics: precision, recall, AUC, subset accuracy. See evaluation results above.

Results

See evaluation results in model card metadata above and metadata.json for detailed per-concept breakdown.

Summary

Model Examination [optional]

[More Information Needed]

Environmental Impact

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

• Hardware Type: [More Information Needed]
• Hours used: [More Information Needed]
• Cloud Provider: [More Information Needed]
• Compute Region: [More Information Needed]
• Carbon Emitted: [More Information Needed]

Technical Specifications [optional]

Model Architecture and Objective

sklearn LogisticRegression (C=1.0) with OneVsRestClassifier wrapper for multi-label classification.

Compute Infrastructure

CPU-based training (8 cores typical)

Hardware

[More Information Needed]

Software

[More Information Needed]

Citation [optional]

BibTeX:

[More Information Needed]

APA:

[More Information Needed]

Glossary [optional]

[More Information Needed]

More Information [optional]

[More Information Needed]

Model Card Authors [optional]

[More Information Needed]

Model Card Contact

https://github.com/pilipolio/chess-sandbox/issues