Ioannis Papathanasiou

Physics graduate · Early-career computational physicist

From dense matter to compact-star observables.

I study how descriptions of dense matter become macroscopic signatures in compact stars, using computational and statistical methods chosen to serve the physics.

Current focus: controlled sound-speed deformations of analytical BSk24, with thermodynamic admissibility assessed before one-fluid reconstruction and stellar calculations.

View selected work Email me
Approach
Question → model → test
Domain
Dense matter → compact stars
Based in
Thessaloniki, Greece
Portrait of Ioannis Papathanasiou overlooking a city at night
Ioannis PapathanasiouThessaloniki · Greece

Research through-line

Trace one change across scales.

My current work asks how a controlled change in a dense-matter model survives thermodynamic assessment and appears in macroscopic compact-star observables.

Four-panel scientific figure defining the BSk24 deformation geometry: a Gaussian is introduced through a smooth window, signed amplitudes change the squared sound speed, and representative raw profiles remain bounded by the causal limit.
Realized deformation geometry for the fixed BSk24 campaign. This is a controlled effective deformation, not a microscopic composition model. Figure: Ioannis Papathanasiou, campaign data v1.0.0, CC BY 4.0; web-optimized from the released PDF with scientific content unchanged.

Current method · analytical BSk24

A controlled perturbation with an auditable path.

  1. Define

    Introduce a local signed sound-speed change above a retained anchor.

  2. Assess

    Reject inadmissible raw proposals before any reconstruction or stellar work.

  3. Propagate

    Carry accepted cases into one-fluid thermodynamics, structure, and tides.

See the complete evidence trail

Selected work

Current work and methodological foundations.

The BSk24 study now leads with its sole-authored arXiv preprint, released software, and campaign evidence. The thesis case study remains distinct and shows how a later methodological review narrowed the defensible classification claim.

Research and reproducibility
2026–present

Signed BSk24 sound-speed deformations

A controlled study of how positive and negative local sound-speed changes propagate through thermodynamic reconstruction and into fixed-mass neutron-star structure and tidal response.

Sole-authored arXiv preprint · software and campaign data public

  • Dense matter
  • Thermodynamics
  • TOV and tides
Read case study

Undergraduate thesis
2026

Compact-star model discrimination

An undergraduate synthetic study of hadronic-star and self-bound strange-quark-star model discrimination, followed by a stricter audit of what the classifier can support.

Thesis examined in 2026; distinct post-thesis redevelopment documented

  • Synthetic data
  • Stellar structure
  • Machine learning
Read case study

Interactive demonstrations

Explore the restricted classifier workflows.

Two independent post-thesis applications make retained synthetic model scores interactive. Their assumptions and scientific interpretation boundaries remain visible before launch.

Baseline demonstration

APR-1-surrogate versus fixed-CFL4 scores

Explore selected mass, radius, and tidal inputs inside one restricted synthetic comparison—not an observational composition inference.

View scope and launch

Perturbation sensitivity

A declared synthetic-error test

Inspect score sensitivity after synthetic perturbation, with compactness excluded from the classifier features.

View scope and launch

How I work

The tool follows the question.

A result is more convincing when its assumptions, numerical behaviour, and limits can be inspected. Reproducibility and validation are built into the method from the start.

Start with the physics

Define the system, relevant scales, and scientific question before choosing the machinery.

Make the model explicit

Expose assumptions, domains, anchors, and admissibility conditions.

Choose the tool

Use numerical, statistical, or machine-learning methods only where they resolve the physical question.

Validate the claim

Test sensitivity, preserve provenance, retain failures, and state the interpretation boundary.

Education

Physics, AUTh

Degree in Physics, Aristotle University of Thessaloniki, 2026 · 240 ECTS · 8.46/10 (“Very Good”). Undergraduate thesis: 10/10.

Background and approach Public outputs