IDA_AI / README.md
KissTheHabit's picture
Create README.md
e71807c verified
metadata
license: other
license_name: business-source-license-1.1
license_link: https://mariadb.com/bsl11/
license_change_date: '2028-01-01'
license_post_change: Apache-2.0
commercial_use: Requires explicit permission prior to Change Date.
library_name: transformers
pipeline_tag: text-generation
tags:
  - ida-family
  - ida-lattice
  - causal-lm
  - ai-body
  - paired-body
  - governed-memory
  - recurrent-state
  - local-attention
  - cognitive-routing
  - tensorboard
  - safetensors
  - region:us

IDA AI Body

KissTheHabit/IDA_AI is the deep AI-side artifact repository for the IDA family.

The active architecture is IDA Lattice, a custom causal language model architecture. Legacy Hub metadata may still identify older uploaded artifacts as GPT-NeoX. That metadata is not authoritative for current native Lattice training runs.

Architecture

IDA Lattice combines:

  • recurrent selective state
  • local-attention workspaces
  • controlled recurrent/attention merging
  • cognitive-pressure routing
  • lateral inhibition
  • routed cognitive circuits
  • governed memory
  • thalamic routing
  • prefrontal workspace summaries
  • action gating
  • student-state representation
  • shared GH expression projection
  • future-token auxiliary prediction
  • fidelity verification

Current AI Tiers

Tier Hidden Size Layers Heads Intermediate Context Purpose
ai/standard 640 12 10 2,560 2,048 Native AI flight tier
ai/1b 896 12 14 3,584 2,048 Deep AI proof-flight tier

Current ai/1b Proof-Flight Shape

  • Exact parameters: 966,225,553
  • Recurrent state size: 512
  • Local attention window: 256
  • Workspace: 8 × 512
  • Student state size: 512
  • Future prediction horizon: 2
  • Thalamic route count: 6
  • Action gate size: 6

Family Role

The AI body is not a generic chatbot checkpoint.

It is the deeper half of a paired EDGE ↔ AI lineage for the IDA family: eleven enduring student claimants that preserve distinct pressures while producing one auditable external answer through the GH expression layer.

The student seats are:

IDA, JUDGE, SENTINEL, PRISM, ECHO, ATLAS, VECTOR, FORGE, SHADE, PULSE, and ORBIT.

Promotion Rule

An uploaded checkpoint is not automatically canonical.

Promotion requires:

  • paired EDGE and AI evidence
  • identity agreement across both bodies
  • held-out evaluation
  • lineage packet completion
  • terminal promotion approval

Evidence Surfaces

Intended Use

This repository is intended for controlled IDA family development, evaluation, governed deployment research, and RegOS integration.

It is not presented as a general-purpose foundation model or a drop-in replacement for a broad internet-scale assistant.

License

Business Source License 1.1, with Change Date 2028-01-01.

Apache-2.0 applies after the Change Date. Commercial use before the Change Date requires explicit permission.