Self-Effect Annotation Layer¶
Annotation layer — instantiates Self-Generated Signal Cancellation
Tags the self-caused component in the stream instead of deleting it, preserving the predicted self-effect and the residual side by side for audit.
Sometimes you must not throw the self-caused signal away — you have to keep it, labeled, so a human or a downstream process can see both what the system did and what the world added. Self-effect annotation layer is the mechanism that, instead of subtracting the predicted self-effect, marks it: it overlays the observation stream with tags that say "this stretch is self-caused, this much, with this confidence — and this leftover is not," leaving the original data intact. Its defining move is non-destructive labeling for audit: where subtraction removes and a gate escalates, this layer preserves, annotating the predicted self-effect and the residual together so attribution is visible and reviewable rather than silently applied. It is the mechanism you reach for when the record itself matters and someone will need to check the attribution later.
Example¶
A large asset manager's compliance team watches for market manipulation, but the firm's own trading is itself one of the biggest movers of the prices it monitors — a block order to sell 200,000 shares pushes the price down all by itself. If the surveillance system deleted the firm's own footprint, it would lose the very data an auditor later needs; if it ignored it, the firm's own selling would look like a suspicious downdraft. The self-effect annotation layer takes the third path: as each of the firm's orders executes, the tape is annotated — this price move tagged "self-caused, attributable to our order of this size at this time, expected impact this many basis points," while the unexplained residual move is tagged "external." Nothing is removed. When a regulator or an internal reviewer examines the day, they see the full price history with the firm's own market impact labeled and separated from genuine market movement, and can audit whether the attribution was fair. The layer never decides guilt and never cancels a print — it makes the self-versus-world split legible on a record that must survive review.
How it works¶
- Mark, don't remove. The predicted self-effect is written into the stream as metadata over the original samples, which stay intact.
- Carry both sides. Each annotation pairs the predicted self-caused component with the residual, plus a confidence, so a reviewer sees the whole attribution, not just a verdict.
- Preserve provenance. Tags record which command and which prediction produced the self-effect estimate, so the attribution can be re-checked later.
- Defer the decision. It labels; acting on the label is left to a human or a downstream stage.
Tuning parameters¶
- Tag granularity — per-sample marks versus coarse episode tags; finer supports precise audit but bloats the record.
- Confidence disclosure — whether each tag carries an uncertainty; showing it aids honest review but complicates automated consumers.
- Retention depth — how long annotated streams are kept; longer serves audit but costs storage.
- Overlay vs. sidecar — whether tags live inline or in a parallel channel; inline is self-contained, a sidecar keeps the raw stream pristine.
When it helps, and when it misleads¶
Its strength is that it preserves information a subtractor would destroy: the labeled record supports dispute, audit, and after-the-fact re-attribution, which is why regulated and safety-critical domains often prefer tagging to silent removal. It keeps data provenance — the traceable record of where each part of a signal came from — intact.[n1]
It misleads when the labels are trusted as if they were ground truth: an annotation is only as good as the prediction behind it, and a confidently mislabeled self-effect can excuse a real event as "just us." The classic misuse is over-tagging — marking so much as self-caused that a genuine external signal is quietly written off in the metadata. The guarding discipline is to carry each tag's confidence with it and to keep the raw, un-annotated stream recoverable, so a reviewer can always re-run the attribution rather than inheriting it.
How it implements the components¶
predicted_self_effect_profile— it marks the predicted self-caused component in the stream rather than removing it, using the profile as the content of the label.self_effect_audit_trace— it produces the reviewable record that exposes predicted self-effect and residual together for human or automated audit.
It does not delete the self-effect to leave a bare residual (residual_difference_signal — Predicted-Signal Subtraction), decide when a residual is a real event (residual_attribution_rule — Residual Mismatch Gate), or route the command copy in the first place (command_copy_channel — Motor-Command Echo Routing).
Related¶
- Instantiates: Self-Generated Signal Cancellation — it is the non-destructive branch of the pattern, marking rather than removing the self-effect.
- Consumes: Forward-Model Prediction supplies the predicted self-effect it labels.
- Sibling mechanisms: Corollary Discharge Pathway · Motor-Command Echo Routing · Forward-Model Prediction · Predicted-Signal Subtraction · Residual Mismatch Gate · Forward-Model Recalibration Loop
Editorial Notes¶
Form Classification¶
Form family: Record, Log & Register
Rationale: Self Effect Annotation Layer operates by writes predicted self-effects, residuals, confidence, and provenance into the original sample stream. That concrete deployed or enacted form is Record, Log & Register under the frozen taxonomy.
Nearest alternative: Interface, Display & Cue — Although Interface, Display & Cue can support this mechanism, the frozen evidence makes its operative form the act that writes predicted self-effects, residuals, confidence, and provenance into the original sample stream; the alternative is therefore secondary rather than defining.
Review outcome: Adjudicated after independent review; high confidence.
Origin Attribution¶
Primary origin: Systems Thinking & Cybernetics
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Marking which observations are expected consequences of the system's own actions instantiates the reafference/efference-copy distinction in feedback systems. NIMH and NIH sources describe action-perception loops that predict and subtract self-generated effects; engineering transfers that distinction into annotated telemetry.
Related originating lineages:
- Data Science & Analytics — Layered provenance annotations preserve transformed and residual signals for later analysis.
- Engineering & Design — Instrumentation practice records compensation terms rather than silently erasing them, preserving auditability.
- Neuroscience — neuroscience contributes self-generated sensory effects, neural prediction, and action-perception loops to this mechanism's defining operation—Tags the self-caused component in the stream instead of deleting it, preserving the predicted self-effect and the residual side by side for audit—without displacing the selected primary historical lineage.
- Robotics & Automation — Efference-copy and forward-model control explicitly predict self-caused sensory change.
- Statistics & Experimental Design — statistics_experimental_design contributes measurement, perturbation design, inference, and uncertainty analysis to this mechanism's defining operation—Tags the self-caused component in the stream instead of deleting it, preserving the predicted self-effect and the residual side by side for audit—without displacing the selected primary historical lineage.
Review resolution: The blind reviewers disagree on primary lineage (systems_cybernetics versus engineering_design). Authoritative or primary research supports systems_cybernetics as the best historical origin: Marking which observations are expected consequences of the system's own actions instantiates the reafference/efference-copy distinction in feedback systems. NIMH and NIH sources describe action-perception loops that predict and subtract self-generated effects; engineering transfers that distinction into annotated telemetry. The cited NIH/NIMH, Research Domain Criteria: Perception and Action-Perception Loops; NIH/PubMed Central, Efference Copy in Kinesthetic Perception directly supports the mechanism's defining operation. All independently supported contributing domains are retained without an arbitrary cap. origin_mode=cross_disciplinary_synthesis records the lineage relationship, while domain_reach=multi_domain records later applicability separately from provenance.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
- NIH/NIMH, Research Domain Criteria: Perception and Action-Perception Loops
- NIH/PubMed Central, Efference Copy in Kinesthetic Perception
Notes¶
[n1] Data provenance — the documented record of a datum's origin and the transformations applied to it. Annotation preserves provenance by labeling the self-caused contribution in place rather than erasing it, so downstream users can trace and re-check the attribution. ↩