Understanding life, more deeply.

Vivethos is building reproducible computational methods for biology. Our standard: register each claim before testing it, check it against controls and independent data, and publish its limits, including when it fails.

Vivethos is a private, early-stage initiative run by a single founder. It has begun computational analyses on public data; none is published yet. It has no publications, no wet-laboratory or clinical work and no validated biological findings. The background is a decorative, AI-generated conceptual illustration and carries no data.

Where things stand today

We publish our status, not only our ambitions.

Research programs
2, computational
Data
Public, openly licensed
Publications
None yet
Laboratory or clinical work
None

Research programs

Two computational programs, each stated as a question that evidence could answer either way.

R-001

Adaptive Biological Discovery

When does choosing the next genetic perturbation experiment with a model beat choosing at random, and how should that be measured fairly?

Computational study of how to select informative genetic perturbation experiments, and of which evaluation methods give trustworthy answers.

Computational analysis
In progress
Protocol public
Not run
Independent replication
Not run
Experimental validation
Not started

Simulation on existing data is not a prospective experiment. Nothing here has been validated in a laboratory.

Conceptual illustration: a network of points on a deep navy field; some points are chosen and ringed with dashed circles, others are marked by hollow squares.
Conceptual illustration, AI-generated. Dashed rings mark experiments chosen by a strategy; squares mark a held-out reference panel. No data are shown.

R-002

Virtual Tissue Laboratory

What can a multicellular simulation honestly tell us about tissue organisation, and which of its predictions can be tested against measurements?

Mechanistic multicellular simulation with numerical verification, and a strict separation between a model that runs correctly and a model that is biologically valid.

Numerical verification
In progress
Biological validation
Blocked
Protocol public
Not run
Experimental validation
Not started

A simulation that reproduces a pattern does not show the pattern has that cause. Biological validity is not claimed.

Conceptual illustration: a packed mosaic of translucent cells in navy and teal tones forming flowing domains.
Conceptual illustration, AI-generated, of a simulated tissue. It is not simulation output and not a microscope image.

Latest progress

What was done, with its status. Holds and corrections included.

  1. 17-18 Nov 2026

    Adaptive Biological Discovery (R-001): analysis workflow built and quality-checked

    A written protocol, a reproducible computational workflow on public perturbation screens, and negative controls were built. An internal quality review led to corrections, which were recorded openly, and a replication scope was drafted. No result is public yet.

    In progress
  2. 17-18 Nov 2026

    Perturbation prediction baselines checked for leakage

    Simple baseline predictors were compared with a published benchmark model under leakage checks. A reproduction defect and a script error were found, disclosed and fixed. Further model training is on hold by pre-set rule.

    In progress
  3. 17-18 Nov 2026

    Biomedical foundation-model evaluation: gates frozen, one error corrected

    A preregistered evaluation was frozen before inference. An earlier claim about possible training-data overlap was found to be wrong and was formally corrected. Further benchmarking is on hold.

    Held
  4. 17-18 Nov 2026

    Virtual Tissue Laboratory (R-002): simulation environment verified; biological validation blocked

    The simulation software environment was integrity-checked and its run times measured. An eligibility review concluded that the current model cannot yet be tested against a matching public experiment, so the main bottleneck is now biological testability, not software. This is recorded as blocked, not as a result.

    Blocked

Our standard

Six steps, in order. This is the process we hold ourselves to; the ledger will show where we meet it and where we do not.

  1. Define the question

    A precise, falsifiable question, with a statement of what result would make us stop.

  2. Check prior art

    Targeted review of what is already known, so that replication is not presented as discovery.

  3. Register the plan

    Datasets, baselines, controls, metrics and decision rules are written down before outcomes are examined. Public registration begins with the first ledger entry.

  4. Test against controls

    Held-out and independent data, negative controls, and the strongest simple baseline, not a weak one.

  5. Seek independent review

    Adversarial review by someone other than the builder. Vivethos is one person today, so none has been publicly documented yet. Agreement is not proof.

  6. Record the result

    Every claim enters the evidence ledger with its status, its limits and its history, including negative results.

The evidence ledger

A public, append-only record of every claim Vivethos makes: its status, what would falsify it, and what was found. The format is shown below with placeholder rows.

EntryClaimStatusWhat would falsify it
VIV-EX-001Example: a strategy for choosing experiments outperforms random choice on independent data.Proposed No improvement over random on held-out independent data at the preregistered budget.
VIV-EX-002Example: a result measured on the same experiment it was designed on overstates performance.Reported Scores on independent data match scores on the original within the preregistered tolerance.
VIV-EX-003Example: a simulated tissue pattern matches a published quantitative experiment.Blocked Simulation falls outside the preregistered tolerance of the published measurement.

These rows demonstrate the format. They are placeholders, not findings. The live ledger will open with the first registered protocol.

Where we are looking

The areas our two programs touch. Interest, not capability.

  • Genetics and genomics
  • Molecular and cellular biology
  • Computational biology
  • Biological simulation

Not yet a record. A standard.

Until there are results to show, we show how we will decide what to believe.