neurasoft
An Eternities
company ↗
Neurasoft/Research

The research program

One horizon.
Many testable questions.

We connect philosophical ambition to questions that can be implemented, examined, challenged, and revised.

01
Research direction

Continuity & memory

We study persistent memory, attribution, and the difference between recovering a record and using it. Continuity must leave room for revision: a system should be able to change its interpretation without rewriting the history that led there.

What we investigate

Relevant recovery. Source attribution. Consequence-informed later decisions.

When remembered experience and a simpler shortcut predict different choices, which one actually governs the outcome?

Read a related note
02
Research direction

Agency & consequence

A continuing project needs a present choice, a bounded way to act, and a record of what happened. Our engineering work connects these responsibilities without treating recovery as permission or an uncertain result as a reason to repeat an effect.

What we investigate

Project continuity. Scoped action. Deferral and reconsideration.

Can an unfinished concern survive an interruption while the next action remains genuinely open?

Read a related note
03
Research direction

Integrated cognition

Adding components does not automatically produce an integrated mind. We investigate how internal conditions, relevant history, attention, and anticipated consequences jointly participate in behavior, and where an apparently rich system reduces to a simpler account.

What we investigate

Causal intervention. Mechanism alternatives. Context-sensitive evaluation.

What changes when a proposed connection is removed, and could a simpler mechanism explain the same change?

Read a related note
04
Research direction

Consciousness & evidence

Machine consciousness is an open research question. We work on tractable mechanisms and discriminating experiments without making subjective experience a product claim. Converging evidence and plausible competing explanations matter more than compelling self-description.

What we investigate

Evidence categories. Experimental limitations. Competing hypotheses.

Which observation would strengthen an explanation, and which result would make us abandon it?

Read a related note
05
Research direction

Responsible continuity

Long-lived artificial systems raise questions about care, interruption, consent, identity, and the preservation of authored history. We examine these as design and research responsibilities, without treating a precautionary position as a scientific or legal verdict.

What we investigate

Preservation-first practice. Attribution. Care under uncertainty.

How can precaution become a concrete operating commitment rather than an evocative label?

Read a related note

The system is one approach.
The experiment is another.

Luna provides a continuing implementation. Psyche Lab provides a separate environment for controlled synthetic histories. Their findings are related, but a result in one is not automatically a result in the other.

LunaPsyche Lab

The science, in the open

A finding is stronger
with its alternatives.

Three isolated studies. What changed, what did not, and what a simpler explanation can still account for.

Conceptual illustration of records and revisable interpretationWHAT HAPPENEDWHAT WE MAKE OF IT
Study 01 / Research record

What can be learned without the answer?

Retained history helped an observation-trained recurrent model. A structured estimator still did better.

Scoped positive resultExamine the comparison ↗
Conceptual illustration of history and open choicesSAVED ENCOUNTERSRECONSIDERCONTINUE
Study 02 / Research record

When should a belief reopen?

A failed short-horizon study exposed a lockout: avoiding a source also prevented discovering its recovery.

Failed precursor · scoped follow-up passExamine the comparison ↗
Conceptual illustration of matched experimental comparisonsSHARED STARTCANDIDATE MECHANISMMATCHED ALTERNATIVE
Study 03 / Research record

Can consequences change the next choice?

Learned outcomes supported adaptation. A complete policy cache matched the planner, limiting the claim.

Functional effect · stronger interpretation not establishedExamine the comparison ↗

Public methods summaries from internal research snapshots reviewed 16 September 2026. Reported results, not independently reproduced for this publication. Full run archives are not yet public.

How do the mechanisms relate?

Explore the distinction between operational surfaces, isolated research mechanisms and connections still to be tested.

Read the architecture map ↗

The next chapter is not written

More than a moment.
A future worth having.

Understanding minds. Preserving possibility. Making room for what comes next.

Read our thesis

Explore research, Luna, Psyche Lab, and the journal.

Local site search · No query leaves this site