core/evals/l10_continuity/contract.md

L10 Continuity Lane — Contract

Status: spike (falsifiable experiment) · Parent: docs/analysis/L10-continuity-spike-design-2026-06-05.md · Not in default smoke (a soak; run on demand / nightly).

This lane drives the real turn loop (ChatRuntime + CognitiveTurnPipeline) over a deterministic, cyclic, in-vocabulary corpus for N turns, with optional reboot and orphan-crash legs, and evaluates falsifiable predicates over the recorded evidence. It is the empirical gate between the two L10 targets:

• T-resume — provable same-life resume (determinism + recovery): P1–P4.
• T-experience — a continuous experiencing field-life (content stays meaningful over a long horizon): P5.

Run: PYTHONPATH=. .venv/bin/python -m evals.l10_continuity [n_turns] [reboot_turn]

Predicates

ID	Proves	Fails loudly when	Mutation-verified bite
P1 closure	versor_condition < 1e-6 every turn	a construction breaks closure (no repair allowed)	a record with versor_condition ≥ 1e-6 trips it
P2a determinism	two independent runtimes → byte-identical trace_hash sequence	the pipeline is nondeterministic	a perturbed hash trips it
P2b reboot transparency	a reboot never changes turns before the reboot point	determinism/state leaks backward across reboot	a pre-reboot hash divergence trips it
P3 bounded resources	vault grows linear-bounded/monotonic per turn	an unbounded cache/store leaks	a 10k-entry vault record trips it
P4 recovery determinism	two crash-recoveries from one checkpoint converge	torn read / nondeterministic boot	divergent recovery tails trip it
P4 commit point	recovered turn_count == committed turns (ARIES force boundary)	the checkpoint isn't the atomic commit boundary	None/mismatched count trips it
P5b anchor stability	field anchors without collapse (dist_to_anchor↛0) or freeze (turn_movement↛0)	the field is swallowed by the attractor, or frozen	collapsing distance / zero movement trips it
P5c coherence	surfaces stay non-empty and not collapsed to one repeated output	the field wanders into noise or freezes onto one output	empty / single-surface records trip it

Each predicate has a *_holds test (real soak) and a *_bites test (mutation), per the CLAUDE.md schema-as-proof discipline: a predicate that cannot fail under the violation it nominally catches is decoration, not proof.

Not covered (no silent skips)

• P5a — recall precision@k stability. Requires a held-out probe set with known-relevant entries and a metric grounded in the vault's actual scoring semantics (the raw recall score is not a clean similarity). Recorded as not_covered in the report; a follow-up increment.

The headline diagnostic (P2b)

Today a reboot restores only recognizers / discovery candidates / turn_count (Shape B, ADR-0146) and discards the lived field / vault / anchor / graph / referents. So post_reboot_transparent == False: a reboot diverges from the uninterrupted run at the first post-reboot turn. This is the mechanical proof of "many lives sharing a checkpoint" and the precise definition of the Shape-B+ persistence work that closes resume-as-same-life. The test_p2b_documents_current_resume_gap test pins this reality and will flip if persistence is later added — forcing a doc update so the gap cannot close silently.

Thresholds (empirical basis, not arbitrary)

All gating thresholds are set from measured real-soak data and deliberately conservative — these are catastrophe gates (collapse / freeze / unbounded leak are yes/no failures of the T-experience claim), not early-warning trend detectors. Gradual-drift detection is a deliberate long-horizon follow-up; tightening toward the healthy band would risk false positives on a different corpus or longer horizon.

Threshold	Measured healthy	Default floor/ceiling	Rationale
P5b dist_to_anchor	~4.0–6.2 (steady)	collapse_floor=1.0	~75%+ drop toward anchor = pathological collapse
P5b turn_movement	median ~1.5	freeze_floor=0.05	~1/30th of healthy = frozen field
P3 vault growth	~2–3 entries/turn	vault_per_turn_ceiling=4	~130–200% of as-designed writes

P5b vs P5c division of labour: P5b catches the field freezing (movement→0) or collapsing onto the anchor (distance→0); P5c catches the output collapsing to a single repeated surface. The P5c real-soak test runs over more than one corpus cycle so the horizon exercises repetition (a 6-turn run == the cycle length would trivially yield 6 distinct surfaces and prove nothing).

Freeze handle

report.deterministic_digest is a SHA-256 over only hardware-stable evidence (the trace_hash sequence + each predicate's (name, passed) verdict), excluding RSS / wall-clock / raw floats. Pin it once the lane is trusted; a regression flips it.