Discord

Checklist NEOAI 2026 Homework (take-home practice problems) · Task 2

Terminal Animals

Classify the breed of a cat or dog from a 128×128-character ASCII-art rendering of its photograph.

  • Vision
  • NLP
  • Multi-class classification of ASCII images
Read the official statement

The task

Each object is a former photograph converted into a string of 128 lines of 128 characters (lines separated by "\n"); characters of varying density convey lighting, shadows and contours. The training set has known breed labels and the test set must be labelled.

Abridged by SOTA from the official materials. The official statement has the exact rules, and it wins wherever this summary differs.

At a glance

You get
train.csv (img: ASCII string, label: class) and test.csv (id, img); sample_submission.csv (id, label).
You submit
CSV with columns id, label.
Scoring
Micro-averaged F1 score.
Rules
  • External data are strictly prohibited.
  • Pretrained weights may be used only for google-bert/bert-base-uncased and gpt2; every other model or architecture (including CatBoost, TF-IDF and any predefined architecture) must be initialised randomly and trained from scratch on the provided data.
Format
Take-home practice problem for NEOAI 2026 participants (Kaggle, 12 Apr – 1 May 2026). Individual.

Notebooks

Competition

Solution

Details

Year
2026, Online
Round
Homework (take-home practice problems) · Task 2
Language
English
License
Per-competition Kaggle data licence: Subject to Competition Rules (Code Stars, Find Comments); MIT (Terminal Animals), as stated by the source