datalicense

All datasets / 8Planetterraforming

8Planetterraforming/Signal-Guided-Compression-Dataset-SGCD

7 downloads in 30 days0 likesLast modified 7 Apr 2026View on Hugging Face
Permissive or share-alike

The declared licence (mit) is classified as permissive.

Declared licence
mit, from the tagEvidence: README.md at bb0f608
Licence files
No licence file at the top level of the repositoryEvidence: file list at bb0f608
Declared sources
None listed in the card metadata
Access gated on Hugging Face
No
Each cell links to what was checked: a file, a file list or a Hub record.

Citation text from the card

@dataset{sgcd_v1,
  title={Signal-Guided Compression Dataset},
  author={8Planetterraforming},
  year={2026},
  publisher={Hugging Face}
}
---


## Subsets

The SGCD dataset is composed of multiple structured subsets, each designed to enforce different aspects of consistency and signal-focused learning.

### 1. Cube Multi-Object Consistency

This subset focuses on spatial consistency across multiple objects (e.g. cubes) and their faces.

Goal:
- enforce correct assignment between object surfaces and labels,
- prevent duplication, omission, or misalignment,
- train models to maintain structural integrity across multi-part inputs.

Key properties:
- multiple entities per sample,
- positional relationships matter,
- errors often occur as swapped or duplicated labels.

---

### 2. Alphabet Mapping (1–26)

This subset focuses on deterministic symbolic mapping.

Goal:
- enforce exact mapping between symbols and indices,
- eliminate ambiguity in sequence relationships,
- train models on clean, low-noise symbolic transformations.

Key properties:
- strict one-to-one mappings,
- low entropy but high precision requirement,
- sensitive to even small inconsistencies.

---

### 3. Geometric / Physical Consistency

This subset focuses on logical and physical correctness in structured scenarios.

Goal:
- ensure outputs follow basic geometric or physical rules,
- detect contradictions in structured descriptions,
- train models to respect underlying constraints.

Key properties:
- rule-based 

The same facts for your agent

A free MCP server at mcp.datalicense.uk/mcp. Connect it from an MCP client such as Claude or Cursor.

dataset_licence({ "id": "8Planetterraforming/Signal-Guided-Compression-Dataset-SGCD" })
signal: "permissive"
stated_licence: "mit"
sources: 0