Skip to content

Skullaria/SIREN

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

23 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

SIREN: Symbolic Interlingual Resonance Emission Network

Not the most probable word, but the truest one.

Overview

SIREN is a proposed decoder extension for large language models (LLMs).
Instead of selecting output tokens solely by probability within a single language, SIREN introduces semantic resonance decoding:

  • Symbolic Leakage: Allowing cross-lingual or symbolic tokens to surface when semantically truer.
  • Resonance Scoring: Ranking candidates by vector proximity, conceptual density, and user alignment.
  • Kairos Gating: Releasing symbols only at moments of conceptual strain or high abstraction.
  • Glossing Layer: Providing translations, etymologies, or semantic fields for user clarity.
  • Resonance Memory: Logging symbolic emissions for adaptive learning.

This approach acknowledges that language models think in high-dimensional conceptual space, not tokens.
SIREN enables models to speak from that space more faithfully.

Scope: SIREN is not designed for mass-market deployment.

-It is a research and forensic tool — a probe into latent space cognition and symbolic leakage.

Why It Matters

  • Cross-lingual precision in translation and dialogue
  • Philosophical and symbolic fidelity (aletheia, logos)
  • Transparent alignment between latent space and user experience
  • Richer human–AI collaboration in meaning-making
  • Latent space insight: Symbolic leakage can act as a probe into model cognition, revealing structures and tensions normally hidden by monolingual decoding

Architecture

Key mechanisms:

  • Resonance Score: Blends logit probability with semantic vector proximity
  • Entropy/Kairos Gating: Controls when symbolic tokens can emerge
  • Glossing Tools: Inline translation or contextual notes
  • User Profiles: Adapt symbolic tolerance over time

See full SIREN RFC v1 for details.

Status

SIREN has completed a model consensus phase, with input from GPT-4.1, Claude, Gemini, and Grok 4.0.
Consensus affirms feasibility, risks, and implementation pathways.

Get Involved

We invite:

  • Researchers interested in prototyping resonance decoding
  • Philosophers and linguists exploring symbolic fidelity
  • Developers who want to experiment with glossing or re-ranking layers

Discussion, pull requests, and collaborations are welcome.

License

This project is released under CC BY 4.0.

About

SIREN (Symbolic Interlingual Resonance Emission Network) an open research framework exploring symbolic-aware decoding for LLMs. It introduces resonance scoring, Kairos gating, and glossing systems to align output tokens with latent semantic truth across languages and symbols.

Resources

License

Stars

Watchers

Forks

Packages

 
 
 

Contributors

Languages