PhaseBridge: A Strict Lossless Mapping Between Discrete Alphabets and Phase Space for Cognitive Systems
Year: 2025
Type: article
Abstract: PhaseBridge is an experimental open-source framework that defines a strict, lossless mapping between discrete alphabets (uintM) and continuous phase space (θ ∈ [0, 2π)). It provides a minimal and verifiable interchange format (PIF) that enables reversible round-trip conversion across data modalities such as time series and images. Unlike compression-oriented containers (Parquet, CBOR) or learned embeddings, PhaseBridge emphasizes reversibility, integrity, and a uniform diagnostic signal. Each mapping is guaranteed bit-exact under the declared alphabet when meta.note = "no_processing". Integrity is enforced via SHA-256 hashes, while κ-coherence (|⟨e^{iθ}⟩|) provides a simple, read-only measure of structural coherence. The framework includes:- A Python SDK and CLI tools (pb-encode, pb-decode, pb-validate, pb-kappa).- A formal JSON Schema for PIF v1, with optional MessagePack serialization.- Conformance tests, including property-based tests for round-trip correctness. Demonstrations include strict round-trip for synthetic time series and grayscale images, with hash and κ validation. The approach complements existing formats by providing a minimal representational substrate that can be embedded into wider analytic or multimodal pipelines. PhaseBridge is intended as a foundational layer for cognitive systems where strict invertibility, integrity checks, and cross-modality consistency are essential.
Authors: Fedotov, Anton, Efimov, Sergey, SynqraTech
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7084608730