gradient

Gradient / asymmetric

Description

The structural property that a dimension has direction — moving one way is cheap, the other expensive; one state attracts, another repels; one outcome is more likely than the symmetric alternative. The gradient form is the anti-symmetry schema: when you see a situation where the symmetric framing feels off, gradient is often the correct framing. Often used as a corrective to “binary” or “symmetric” framings that lose information.

Encounters

• “Asymmetric gate” — browser-eyeball-as-gate doctrine: one direction cheap, the other load-bearing.
• “Context economy vs session quality tips the balance” — agent orchestration: gradient between two competing pressures.
• “The math inverts at 76%+” — threshold observation: gradient becomes a step-function at a specific point.
• “Asymmetric stability for fuzzy identity” — KCC feedback memory: for fuzzy identity, prefer the empirically-consistent choice over the theoretically “right” one. Asymmetric resolution criterion.

When it applies / triggers on

User-initiated: Trigger-verb signal in user messages is essentially absent — gradient is overwhelmingly a re-framing reach on the agent’s side rather than a lexically marked prompt cue. The form fires when the user describes a binary or symmetric framing and the agent recognizes the actual structure is directional, or vice versa. Three recurring sub-shapes:

• Cost-asymmetric direction — cheap forward, expensive backward (this is the asymmetric-gate bundle’s home base). Example: “the browser-eyeball step is genuinely asymmetric — overkill for schema PRs, load-bearing for interaction PRs” with the operational heuristic “does this PR change how something feels in the hand, or just what something contains?”
• Inflection / threshold — a parameter on a gradient hits a step-function (“the math inverts at X%,” “tips the balance”).
• Asymmetric-stability — the user names asymmetric stability (“if one choice gives us more stability… i’d lean that way”); the agent elevates to a doctrine: for genuinely fuzzy questions where there’s no ‘right’ answer, prefer whichever choice the extractor/LLM produces consistently. This sub-shape generalizes across decision domains (identity rules, filter-vs-label, even visual-grammar where symmetric/asymmetric marks have predictable semantic loadings — “boat heading somewhere” vs “stamp on a document”).

Agent-initiated: Engine notices the user is reaching for a binary/symmetric framing of a directional situation; surfaces gradient as the corrective. Of 30 top T2 candidates, 4 are <task-notification> flavor (elephant evaluating a goldfish’s directional cost/risk analysis).

Vocabulary cues: “asymmetric,” “asymmetry,” “gradient,” “directional,” “one direction cheap,” “expensive in the other direction,” “tips the balance,” “math inverts,” “asymmetric stability,” “stability beats correctness,” “ratchet,” “threshold,” “inflection,” “tipping point.”

Situation-shape signals: A decision being framed in binary terms when the underlying structure has a cost or stability differential along a continuous axis; a doctrine being proposed as universal when it should be artifact-class-conditional (cheap-and-overkill for class X, load-bearing for class Y).

Composes with

• asymmetric-gate — the specific bundle where gradient meets a gating mechanism (one direction cheap, the other expensive).
• local-minimum — gradient + attractor; every local direction looks worse, but the global landscape has better attractors.

When it doesn’t apply

• Genuine symmetry — mirror-image systems, undirected graphs; forcing a gradient frame loses information.
• Constant landscape — if there’s no actual variation along the dimension, gradient is degenerate.

Sources

• Differential calculus and physics (gradient as derivative).
• Cognitive-science use as a corrective to false-binary framings.

Canonical exemplars from corpus (T2 2026-05-17)

• Asymmetric stability as identity doctrine (cwd: campconnect, session: idx=4): “For genuinely fuzzy identity questions where there’s no ‘right’ answer, the doctrine should pick whichever choice the extractor/LLM tends to produce consistently. Stability beats correctness when both are unclear.”
• Browser-eyeball asymmetric gate (cwd: campconnect, session: idx=6): “The browser-eyeball step is overkill for schema/data PRs (CI catches everything) but load-bearing for interaction PRs (CI catches nothing). A useful heuristic: ‘does this PR change how something feels in the hand, or just what it contains?‘”
• Asymmetric within-week stability (cwd: campconnect, session: idx=5): “Within-week pattern is a fuzzy identity signal… The same shuffle is fine for a card-level descriptive label but is a stability hazard for a filter axis. Prefer the empirically-stable derived display over the theoretically-correct filter axis.”
• Graphic mark asymmetry signals motion (cwd: campconnect, session: idx=9): “Symmetric marks read as static / anchored / institutional. Asymmetric marks read as in-motion / directional / dynamic. Adding the jib cut changes the mark’s feeling from ‘stamp on a document’ to ‘boat heading somewhere.‘”

Trigger pattern (T2): Gradient surfaces when the user describes a binary or symmetric framing and the agent recognizes the actual structure is directional — a re-framing reach on the agent’s side, not lexically marked in the user’s prompt; essentially absent trigger-verb signal in user messages.