Topics — What a work is about. Nucleation holds whatever reading the owner files here, whether or not an Anthology of the SOTA topic could also hold it; instruction for ML practice belongs to the anthology (ADR-048). The words are seeded from the clusters the reading-time triage of 2026-09-25 found, and grow by decision.
Self-governance — governing oneself by standards that are one’s own — identification and wholeheartedness, planning agency and policies, self-constitution, autonomy and authenticity, and the empirical psychology of self-endorsed motives. One of four unities kept apart (group: philosophy; ADR-024).
30 of 191 CASE documents. Back to the full index.
| # | Title | Summary | Status |
|---|---|---|---|
| CASE-013 | The reluctant writer | A writer sincerely committed to finishing an article while avoidance governs the afternoon; who later blocks distracting websites. The individual-scale case (Sections 1, 3.5). | Active |
| CASE-076 | Twin pattern-keepers: two systems keep the same pattern, one by its own responses and one through an external controller, and only removing the controller tells them apart | A10 §3.3’s paired design, the case behind QUESTION-086. Matching behaviour and persistence, different organization. Removing the controller destroys one system’s pattern and not the other’s. A10 then made it harder: the internal regulator at one scale is supported by other processes at another, so which side of the boundary the controller sits on depends on where the boundary is drawn. | Active |
| CASE-077 | The reciprocal pair: two processes that endorse only each other under recursive delegation, so that each one’s authority returns to it through the other | A2 §5’s counterexample to the ban on self-endorsement: “Suppose D1 endorses D2, and D2 endorses D1.” Under g = (1 − η)b + ηEᵀg, each operator’s authority comes partly from its own endorsement returned after two hops. It came back in A6 as authority that “may circulate”. Used to show that a ban on self-endorsement does not make authority independent of one’s own endorsements. | Active |
| CASE-080 | The skeptic who delegates: a critic that cannot overturn an emerging consensus gives its authority to a process that can generate alternatives | A3 §6’s worked example. Four processes: an integrator and a decision- maker push toward a decision, a skeptic disagrees, and a generator can translate neglected evidence into new hypotheses. The skeptic does not try to win. It endorses the generator, whose hypotheses change the decision landscape. Used to show delegation as a route to one’s goals, and dissent made useful by a role. | Active |
| CASE-081 | Self-maintenance learned for a designer’s reward: a system that maintains itself because it was trained to, and later revises the policy and keeps maintaining itself without further instruction | A10 §6.1’s boundary case, “a system that learns to maintain itself only because an external designer rewards that behavior”, with its question whether the maintenance becomes endogenous “once the system can revise the policy and maintain the associated governance without continued external instruction”. It is the case CLAIM-288 is about, and the place where that objection could be answered. | Active |
| CASE-092 v3 | A minimal coupled meaning–influence model: two hypotheses, two retrieval specialists, a fixed balanced corpus, queries set by the dominant interpretation and endorsement updated from apparent retrieval success | A22 §7 and A23 §XI, a simulation not yet built. It is how QUESTION-100 would be answered and CLAIM-297 tested, and what CLAIM-278 asks of the operator framework: a prediction a simpler model does not make. | Active |
| CASE-099 | The two-hypothesis coupled meaning–influence model linearized at its symmetric state: a meaning bias m and an authority bias a, each stable alone, coupled by authority shaping interpretation (β) and interpretation shaping endorsement (γ) | A14 §2 of the exchange of 2026-10-16. Near the symmetric state of the two-hypothesis, two-specialist model, (m, a) evolves by a 2×2 matrix J = [[α, β], [γ, δ]] plus noise. With nonnegative entries and α, δ < 1, the symmetric state is unstable exactly when βγ > (1 − α)(1 − δ), and the unstable direction mixes meaning and authority. Checked here. A construction about a linearization, not a fitted or simulated result. | Active |
| CASE-108 v4 | Robust governance under internal compromise: four governance architectures over cooperative, honestly dissenting and adversarial specialists, scored on attack success and on what the safeguards cost, with a variant that compromises the evaluators | A41’s proposed study, from the exchange of 2026-10-17, on the owner’s U41 question about insider threats. A multitask learner with three kinds of specialist (cooperative, honest but dissenting, adversarial) is run under unrestricted, reliability-weighted, role-limited and adaptive governance, and scored on lost legitimate capability, suppressed honest minorities, cost of safeguards and repair of governance after failure, as well as on attack success. Its sharpest variant compromises the processes that evaluate the others’ competence, which tests CLAIM-342’s independent scoring. Federated learning is the natural testbed. Not run. | Active |
| CASE-109 | Two people with the same demanding commitment, the same output and the same effort in the work, who differ in how their intrapersonal conflict is settled: one escalates it to deliberate override, the other’s processes accommodate one another | The owner’s case, U45 of the exchange of 2026-10-17, worked by A45 as Person A (“repeated escalation”) and Person B (“local coordination”) on one writing project. Both exert real effort and produce work of equal quality; only the route by which conflict is settled differs. It is the variation CASE-013 and CASE-051 lack: there one writer’s effort varies across afternoons, here effort in the work is held high in both. It is used for CLAIM-358. Stipulated, so it shows what the claim distinguishes, not that people differ this way. | Active |
| CASE-112 v2 | Bistable authority with a tilt: in ẇ = μw − w³ + h, with w the balance of authority between two regimes and h a higher-order volition favouring one, for μ > 0 the operative regime changes only when |h| passes 2μ^{3/2}/(3√3), so a sincere and efficacious higher-order volition can move the balance without moving the regime, and which regime operates depends on the path of h | A8 §4 of the exchange of 2026-10-17, restated as A12’s Proposition 3 and kept in A15. The saddle-node threshold and the hysteresis are derived and rechecked here. A8: “It’s a countermodel showing why endorsement strength, effective governance, and behavioral transition can come apart.” A8’s “endorsement” is the reflective attitude, not the record’s TERM-053 sense, so h is called a higher-order volition. It models authority between regimes; it does not revive will as an attractor (TERM-003, Superseded). Fictional by construction. | Active |
| CASE-119 | Three demands each requiring one coordinate of an outcome to equal 1, with attainable outcomes (1,1,0), (1,0,1) and (0,1,1): every pair is jointly satisfiable and the triple is not, and a new composite policy that can also produce (1,1,1) relieves the frustration by enlarging the attainable set, with the acceptance sets unchanged and no demand suppressed | From the synthesis the owner relayed on 2026-10-18. A compatibility complex with all three edges and no filled triangle, the minimal structurally frustrated collection of CLAIM-347 with three members. When the organization acquires a composite operation that can produce (1,1,1), ℛ(𝒪) ⊊ ℛ(𝒪′) and the complex gains the triangle: relief in CLAIM-347’s sense, by changing what is attainable, not by weakening a demand. The construction says such relief is possible; whether specialization produces it is QUESTION-133. | Active |
| CASE-123 v2 | The regulator and two accounts: a formally independent regulator weighs two parties’ accounts of a risk, one party comes to supply more of the evidence it reads and the experts it trusts, and the credibility that party gains decides whose evidence is read next; with variants for a one-time perturbation, a press as the evaluator, and an objective that gives one property no weight | A20 §2 of the exchange of 2026-10-20, on the owner’s U20 (“reflexive control and regulatory capture”), with A21 §3’s press variant on the owner’s U21 and A26 §4’s evaluative variant. CASE-099’s linearization with institutional roles: m is the lean of the regulator’s reading, a the standing of the party it favours. A20: “This is an application of the formal model, not an established empirical account of any particular regulatory agency.” No regulator, outlet or party is meant. | Active |
| CASE-124 v2 | A maintained legibility lean as a third state of the linearized meaning–authority model: authority moves the lean r of the categories through which evidence is read, and r feeds interpretation or authority, with its own persistence λ | CASE-099’s two-variable linearization, meaning bias m and authority bias a, given a third state r: the lean of a maintained legibility regime, the categories, reporting requirements and evidence channels through which evidence becomes readable. Authority moves r (ε), r moves interpretation (ν) or authority (μ), and r persists at rate λ. Two witnesses, both checked: one in which the two-variable loop is stable and the three-variable one is not, and one in which the system is stable but r holds a lean long after m and a would have decayed without it. | Active |
| CASE-125 v2 | A hidden governance bit: an output that is fresh fair noise whatever a fixed hidden bit K is, until an intervention h makes the output read K, so the output alone is exactly informationally closed under ordinary operation and keeps nothing about the response to h | A16 §3–4 of the exchange of 2026-10-20, from A15 §2’s deterministic form, and A17’s Proposition 1. A fair hidden bit K, fixed in time; under ordinary operation the output Y is fresh fair noise; under the intervention h = 1 it reads K. The output is exactly closed, and with h = 1 drawn with probability ρ its deficiency for the response to h is ρ bits, no better than a constant. It is the first instance with parameters of CLAIM-122’s supervisory test, which the exchange of 2026-10-17 gave only in words. Fictional by construction; K is a mechanism parameter, not an agent’s governance. | Active |
| CASE-128 | The voluntary delegation: a person gives a friend the password to a distracting application during a calm period and is refused it later under temptation, while another person lives under the same restriction imposed without consent | A18’s CASE C in the exchange of 2026-10-22, from the owner’s request for cases on quiet subordinate processes (U18). A restriction on a constituent process can be an exercise of self-governance, not its defeat, so a non-domination test that counts every override as domination is too simple. Not shown: where a legitimate precommitment ends and an unaccountable one begins. | Active |
| CASE-129 | The alarm that learned and the alarm that was disabled: two systems that once warned too often are both quiet in ordinary operation, and only one responds when a new hazard appears | A18’s CASE B in the exchange of 2026-10-22, written at the owner’s request for cases on subordinate processes “quiet because they’ve been conditioned never to object” (U18). A quiet signal may be calibration or suppression, and ordinary operation cannot tell which. Not shown: how often either occurs in persons or institutions. | Active |
| CASE-144 v2 | The evaluator that stops tracking: two populations of specialists with identical capabilities and task streams, one scored against independent outcomes and one by an evaluator that weights evidence agreeing with its previously favoured specialists, after a shift that makes a marginalized specialist uniquely competent | Proposed in the exchange of 2026-10-23, “Framework Overview Request”, at A5 (section 5) with A4’s “corrigibility trap”, after the owner asked which experiments could “kill multiple birds with one stone” (U4) and which questions simulation could answer (U5). One design bears on CASE-092, CASE-108 and CASE-117 at once, and separates four causes of a failure to adapt. A5’s central question: “Under what conditions does an organization become unable to recognize or act upon evidence that its own organization of interpretation and authority has rendered inconsequential?” Not run. | Active |
| CASE-153 v2 | The self-protecting sentry: a hidden world, safe or dangerous, a sentry process that favours sheltering and can block inspection with weight κg, an investigator that favours inspecting, which reveals the state with 90% accuracy at a small cost; train in danger, switch to safety, and compare a Bayesian baseline, a habit baseline, the learned-authority model, an independent audit and a protected investigator, then cross governance feedback, evidence restriction and correction rights in a 2×2×2 design | Proposed in the exchange of 2026-10-23 (workbench entry of 2026-10-24, U17), after the owner asked to “brainstorm some minimal toy models we could apply to simulate our abstraction for experiments. there may also be some readily available options that fit our constraints already.” The smallest model of the protective lock-in in CLAIM-446, built on CASE-092 and CASE-099. At U18 the owner pointed to CASE-124 as the better core, and the exchange moved there (CASE-156). Not run. | Active |
| CASE-154 | One representation, two tasks, one bounded observer: with X₁, X₂ independent fair bits and an observer that reads one coordinate and cannot compute XOR, the raw pair U = (X₁, X₂) gives error 50% on Q_A = X₁ ⊕ X₂ and 0% on Q_B = X₁, and the compressed V = X₁ ⊕ X₂ gives 0% and 50%; a legibility regime can enable one perspective’s distinction while disabling another’s, with no task-independent sense in which it is better | From the exchange of 2026-10-24 (workbench entry of 2026-10-24, U22), after the owner said “let’s work on those connections”: a deduction from CLAIM-199’s XOR example, checked when filed. It ties bounded legibility to competing concerns: a representation that serves one task can destroy another’s for the same observer, so a failure of one perspective under a legibility regime may be a resource limit and not an exclusion. | Active |
| CASE-155 v3 | Authority as a replicator population: for a distribution π of control over processes with mean qualities q, the expected Maynard-Cross update is π_a q_a / v, Bayes’ rule when q is each process’s likelihood; the Maynard dynamic is the Taylor dynamic slowed by 1/v, so the two share their orbits; mean quality climbs at the rate Var_π(q); and when a process’s score rises with its own share, ℓ′ = ℓ + h + c(2π − 1) has an unstable threshold π* = (1 − h/c)/2, above which a process the evidence disfavours captures the organization | Filed at the owner’s direction, 2026-10-24, after asking whether the tools of Soma et al. (THEORY-899) or related work could formalize the record’s abstract governance dynamics. Four results, each checked when filed, two by script. It reads authority as the population vector of a replicator dynamic. Evidence-tracking authority is then Bayesian, and authority-dependent scoring gives lock-in with an explicit correction threshold. | Active |
| CASE-156 v2 | A legibility filter as the ν of the meaning–authority–legibility model: in a safe world where 15% of raw cues are alarms, an admission rule P(admit | y, r) = σ(c + χry) with c = 0 and χr = 2 admits 88.1% of alarms and 11.9% of reassurances, so 56.6% of the admitted record is alarms with the world and the sensors unchanged; at symmetry the mean admitted evidence is χσ′(c)r to first order, so ν = η_m χ σ′(c); and r_t = λᵗr₀ + ε Σ λ^(t−1−j) a_j is an exponentially weighted history of authority | From the exchange of 2026-10-24 (workbench entry of 2026-10-24, U18–U19). The owner: “i think we already have a three variable system somewhere, where the third component is described as legibility”, then “flesh this out further”. A microfoundation for one of CASE-124’s abstract couplings: selective admission of evidence, checked when filed. It gives r an observational semantics and reads r as the memory of past authority. Selection-aware inference is the control that the pathology has to beat. | Active |
| CASE-157 v2 | Boosting the vote: in the owner’s 2014 model, voters of fixed accuracy drawn from Beta(2, 2) vote in a weight-averaged election, and every voter’s weight is multiplied by 1 + f when their own vote was right and 1 − f when not, with a fraction replaced each round at weight 1; with full-information feedback the weighted result rises from 0.50 toward 0.98 as f grows, and churn only costs in a world that never changes; the model escapes the hive mind’s absorbing states because every voter is scored, heeded or not, and it assumes away the capture its author predicted | Filed at the owner’s direction, 2026-10-24, after the owner suggested taking inspiration from the experiments in dmarx/BoostTheVote. A thought-experiment written in R in 2014, read at commit b5bca8f and re-run here through a Python port (scratch/formal/boost_the_vote.py). It is the full-information counterpart of the replicator construction. What it shows, what it assumes, and the three-axis extension it suggests are recorded, and the extension was run on 2026-10-25. | Active |
| CASE-158 v2 | Audited governance against a contextual router after a reversal of expertise: specialist denoisers with frozen, separately trained operators propose for one shared latent at each step, and spatial grants are revised from independent held-out probes of every proposal, available only from the next episode, against a fixed allocation, a conformity-scored allocation and a contextual mixture-of-experts router; the question is whether governance does anything a router does not, and under full-information audits the record predicts it cannot | Proposed at A46 and reported at A47 of the exchange of 2026-10-24, continued, after the owner’s “close the loop”. It is CASE-144’s design, run without citing it, with competence frozen so that only the allocation learns, and with a conventional router added as the rival that CLAIM-278 requires. A47 reported that the router won on every seed. Its results come from a sandbox the share does not carry, and the record has not run it. What would make it informative is one arm per failure of QUESTION-271. | Active |
| CASE-159 | Authority geometry and response geometry: when an organization’s response is the authority-weighted mixture p_π(y | c) = Σ πᵢ pᵢ(y | c), the Fisher metric of its responses is the pullback JᵀD⁻¹J of the categorical (Shahshahani) metric on the shares π and never exceeds it; it vanishes along any redistribution between processes that respond alike under the probes used, so two specialists who both answer yes with probability 0.8 in routine contexts carry no information about their shares there, while a rare context in which they answer 0.9 and 0.1 carries 0.64/(p(1 − p)) per observation, p = 0.1 + 0.8π | From the continuation of the exchange of 2026-10-24 (U28–U31), where it appears in four replies (A28 §3, A29’s second priority, A30 §6A and A31’s probe diagnostic); checked when filed. The categorical metric on authority shares is the one in which the replicator is a gradient flow (CLAIM-444). What an evaluator sees is the response mixture, whose Fisher metric is that metric pulled back through the processes’ response profiles. Redistribution between processes that agree on the probes is invisible to them, and a probe on which they disagree restores it. The mixture counterpart of CASE-152’s product model. | Active |
| CASE-161 v2 | Reciprocally produced, externally reset and feedback-only governing rules in a 32-member population: with ordinary trajectories identical, reciprocal production repairs the loss of one rule and not of both and fails after producer outages of 36 updates for audit and 40 for admission, after an audit-rule pulse the next audit integrity is 1.00, 0.18 and 0.00 in the three arrangements, and a self-dependent production graph does not regenerate a zeroed rule | Versions 1.6 and 1.7 of the Perspective Population Lab, dmarx/PerspectiveLab@7e2edc2. Two governing rules, the integrity of audit and of evidence admission, are kept up in three ways in the same 32-member population. Ordinary operation cannot tell the three apart, and loss, repair and producer interventions do. It carries CASE-126’s second construction to a stochastic population with a different production law: here reciprocal production repairs a single loss, where the case’s closed pair collapses. | Active |
| CASE-162 v2 | Capture and selectively admitted evidence in two lab realizations: in a four-variable model of authority, record and competence the equilibrium fold under self-scoring is at c ≈ 0.4247 while the declared history is trapped from c ≈ 0.4469, and in a 32-member event-level population, admission of cues that depends on the record reverses the sign of the admitted evidence and leaves the worse alternative in charge in 41 of 64 worlds, which exact inverse-propensity weighting brings to 0 of 64 below full selectivity | Versions 1.1 to 1.3 of the Perspective Population Lab, now held as dmarx/PerspectiveLab@7e2edc2. v1.2 is a deterministic four-variable model whose analytic fold and whose trapping of a declared history come apart. v1.3 is a separately coded event-level population in which the filtering of CASE-156 runs on sampled cues, and a known-propensity evaluator undoes it while admission stays positive. v1.1 adds capture and selective audits to CASE-158’s design, and the router still wins. The sign reversal is CASE-156’s; what is new is two running realizations, with their code. | Active |
| CASE-163 | Identifying how governing rules are maintained from diagnostic probes: a predictor fitted under reciprocal production fails under self-dependent production with histories matched, an audit pulse with a short producer interruption identifies which of four known mechanisms applies and predicts withheld interventions, and fitting decay and threshold as well helps after the parameters shift and hurts at nominal ones, within a known catalog and with every probe run on a counterfactual reset | Versions 2.0 to 2.2 of the Perspective Population Lab, dmarx/PerspectiveLab@9f82bfd, on the maintained population of CASE-161. v2.0 shows that a response model does not transport across a change of production law that ordinary histories do not reveal. v2.1 and v2.2 infer the mechanism, and then its parameters, from probes, and predict interventions never probed. Everything is relative to a catalog of four known mechanisms whose equations the forecaster is given, and every probe is a separate branch from the same starting state. | Proposed |
| CASE-182 | Two operators under one update rule, endorsement moving toward observed outcomes: scored every round, the more accurate one holds the lead durably whatever the start; scored only when it acts, under sharp selection, whichever gained influence early keeps it | Filed by the fourth status review (ADR-057) for CLAIM-277, whose two loops the record had stated and not constructed. Two operators share one rule, with accuracies 0.8 and 0.6. Each round one acts, chosen with probability softmax(β·E), and endorsement moves toward observed outcomes, E ← E + η(outcome − E). Nothing marks either one as an authority. Scored every round, the more accurate one ends more endorsed in 200 of 200 seeds, even after a head start for the other (E0 = 0.3 against 0.9). With η = 0.01 the lead never changes hands in the second half. Scored only when it acts, with that same head start, the worse one is never displaced in 200 of 200 seeds at β = 30 or 100. At β = 8 the better one recovers, after a median of 460 rounds. The scored control at β = 100 entrenches no one. Code: scratch/formal/trust_loop.py. | Active |
| CASE-186 v2 | A proposed study: an organization in a world whose hidden regime reverses and then leaves its hypothesis family, with evidence generation, admission and scoring separately lesionable, its diagnostic machinery maintained by its own reciprocal production, from outside, or not at all, and finite fuel shared among operation, maintenance and diagnosis, to find whether a second correction after injury to the diagnostic and copying machinery separates arrangements that correct the first failure alike | The CASE A22 proposed and A23 wrote out at the owner’s “flesh out the proposed CASE”, in the exchange “Analyze nucleation concepts” (2026-10-31), under A23’s title “Maintained Inquiry Under Interpretive Capture, Productive Damage, and Environmental Novelty”. Its admission × scoring arms repeat CASE-162’s, and its double-lesion prediction is CASE-161’s finding. What is new is the primary endpoint, a second correction after the machinery is injured; an environment that leaves the hypothesis family; and external restoration that is paid for. Not run. The analytic error rates hold only where admission is protected. | Active |
| CASE-188 | A proposed study: two agents keep the same long-term commitment and decide alike under ordinary conditions, one protecting it from irrelevant fluctuation while letting specified defeaters revise it, the other growing ever more efficient at enforcing it while evidence against it loses its way in; varying the environment, the legitimacy of the original commitment and what the evaluative mechanisms can see, to tell corrigible resolve from entrenchment without building the answer into the diagnostic or the score | Proposed by the owner at review of the filing of the exchange “Analyze nucleation concepts” (2026-10-31), as a successor to CASE-186 and kept separate from it so that study stays interpretable. It gives CLAIM-344’s distinction between resolve and mere persistence a counterfactual test. Not run; every parameter is left to a preregistration. | Active |