Argus

Sustaining observation. Continuing investigation.

Argus explores how an intelligent model can sustain its own activity, respond to what it encounters and develop the foundations for continued learning.

Kind
Research programme

The idea behind Argus

Argus investigates an intelligence whose development is driven from within.

We built Argus to investigate how a model could keep observing, recognise when its present approach had reached its limits and continue its investigation.

This brings together several questions. What should a model pay attention to? How should it judge what it finds? How could its developing knowledge influence what it explores next?

Argus studies how observation, discovery and further investigation become connected parts of the model's own development.

Following an investigation further

Early experiments demonstrated a practical part of this work.

Recorded runs demonstrated internally directed continuation: Argus extended its investigation through its own activity.

What was happening within the model helped determine what happened next.

This established an early example of internally directed continuation. It gave us a working foundation for investigating how a model might adapt its attention as it encounters information and develops a more useful understanding of it.

Sustaining observation

Continued investigation depends on the ability to remain active as information arrives.

In a recorded continuous experiment, Argus sustained its observation throughout the run. Its accumulated work remained available for continuation.

This work established a practical foundation for sustained activity. It also brought continuity into the research itself: how a model maintains its operation, preserves its working state and remains able to continue.

For Argus, these questions belong alongside learning. A model that develops through experience needs a dependable basis for that experience to accumulate.

Judging what deserves attention

Discovering something useful also requires careful evaluation of what is found.

Experiments established more dependable measurement of the differences under investigation.

This strengthened the basis for further investigation. It also sharpened an important question for the wider research: what makes an observation worth pursuing?

The value of a discovery depends on the reasons for accepting it.

Developing from within

Argus takes these abilities forward together.

We are investigating how sustained observation can support useful discoveries, how those discoveries can remain available, and how their connections can contribute to future investigation.

Argus is part of our research into intelligence that takes an active role in its own development. The work brings internally directed activity, continuity and the evaluation of what it finds into that investigation.

Continuing to observe. Finding what deserves attention. Developing the understanding that guides what comes next.

Cite this record

Forlais Research (2026). Argus. Forlais Research Library, record FR-P-10.

@misc{fr-p-10,
  author = {{Forlais Research}},
  title = {{Argus}},
  howpublished = {Forlais Research Library, record FR-P-10},
  year = {2026}
}