Technical Research & Engineering

GLOBAL INTENT COMPANY

Interactive technical environment for machine learning, scientific instrumentation, computational systems, experiments, and project evidence.

Developed independently by founder and researcher Cory Tortorici. Source code, controlled experiments, null results, architectural iterations, and public research logs are made inspectable across all project nodes.

Source Repositories

Laboratory Chronology

LATEST FROM THE LAB

Recent empirical findings, architectural iterations, and null results surfaced directly from active research logs.

Active laboratory research log entries with timestamped records are available in the full research log.

Interactive Spatial Knowledge Graph

EXPLORE THE RESEARCH

Navigate the connected network of research projects, architectural pipelines, and domain clusters in real-time 3D.

Rotate • Zoom • Click Nodes to Unfold

60 FPS

Tier:

Focused on CORTEX-MS • Orbiting sub-nodes active

Domains:

SPATIAL FOCUS:CORTEX-MS•Machine Learning

Machine LearningKind: researchActive Research[CORTEX-MS-L4]

CORTEX-MS

Learned molecular reasoning architecture that maintains and updates competing molecular hypotheses from mass-spectrometry evidence rather than one-shot retrieval.

Core Research Question:"Can molecular identification from mass-spectrometry evidence benefit from a learned architecture that maintains and updates competing molecular hypotheses rather than treating identification solely as one-shot classification or retrieval?"

Structural Overview

Cortex-MS transforms raw tandem mass spectra (MS/MS) and precursor metadata into structured geometric peak embeddings. These embeddings are routed into an active workspace where competing hypothesis slots evaluate formula consistency, functional fingerprint predictions, and candidate ranking in parallel.

Design Invariants & Principles

[1]Hypothesis Competition: Rather than collapsing features in a single forward pass, parallel hypothesis slots compete for evidence support.

[2]Geometric Peak Manifold: Peak m/z, relative abundance, and collision energy are projected into continuous metric spaces.

[3]Calibrated Uncertainty: The system preserves candidate margin ambiguities instead of forcing artificial overconfidence.

Sequential Pipeline & Processing Stages

Mass Spectrum InputPrecursor & Collision Context

MS/MS centroid peaks, precursor m/z, charge state, collision energy, and adduct metadata.

Data Contract: Raw peak array [m/z, intensity] + context scalars

Peak EncoderHigh-Dimensional Projection

Fourier-transformed positional encodings mapped to metric vector representations.

Data Contract: Dense peak tokens [N_peaks, 512-dim]

Spectrum EmbeddingGlobal Context Pooling

Multi-head self-attention aggregating fragmentation tree dynamics into a unified spectral signature.

Data Contract: Latent spectrum manifold [1024-dim]

Cortex WorkspaceHypothesis Competition Slots

Three concurrent hypothesis slots (Hypothesis A, B, C) that maintain and update competing molecular hypotheses.

Data Contract: 3x Active Hypothesis States with competition weights

Multi-Task DecodersFormula, Fingerprint & Representation

Joint prediction of elemental formula, Morgan chemical fingerprint bits, and metric contrastive representations.

Data Contract: Formula logits, FP probability vector, Latent embedding

Candidate RankingEvidence Matching

Calculates cosine similarity and margin-loss distances against molecular candidate libraries.

Data Contract: Rank-ordered candidate list with calibrated margins