Reloads every 5 min.

simplify — progress KPIs

progress through the corpus

At a glance

How much of the text is explained, what the rules claim is finished, and how quickly the run is moving.

Corpus coloured

73.0%

52,904 of 72,450 observed clauses have a rule.

Corpus closed

47.8%

34,643 of 72,450 observed clauses are claimed by a rule to need no more processing. This is a rule claim, not proof that the result is true.

Corpus words closed (recurse:leaf)

19.0%

108,532 of 572,631 emitted words are in final output segments explicitly marked recurse: leaf. This is a rule claim, not proof that the result is true.

Shapes with rules

55.1%

15,099 of 27,380 observed powerset shapes are fully covered by rules.

Surprise: sentences until a new pattern

8.8

Source-balanced across 22 corpus sources, capped at 229 recent sentences per source; 90% chance of finding one within 20 sentences. The current single-source stream tail is one per 32.0 sentences.

Corpus size

38,698

sentences in this run

Simplify warm words / second

64.3

Simplify alone on a fixed benchmark after spaCy and the model are loaded.

Simplify warm sentences / second

9.9

Input sentences parsed by Simplify alone on the same warm benchmark.

Full pipeline words / second

38.5

End to end: Simplify, coverage checks, hypertrie, persistence, and publication.

Full pipeline sentences / second

2.7

Completed corpus sentences per wall-clock second across the full diagnostic pipeline.

Corpus processed

100.0%

38,698 of 38,698 sentences processed.

Rules

1,250

rules registered in the measured Simplify engine

Open HITL decisions

61

human choices still waiting in Simplify's review queue

Current run

34.9%13,500 of 38,698 sentences processed
1251rules in the current engine

These are live runner facts. The KPI cards above remain the last complete, stable run.

Is the project moving?

Rules are still being added, but corpus breadth has barely moved: this is depth, not broad coverage.

Last completed engines: 1,249 → 1,250 rules; corpus text explained 73.0% → 73.0%. The current run is 13,500 of 38,698 sentences (34.9%) and reports 1251 rules; its partial percentages are not compared with the completed-run baseline.

Board signal: rules and corpus coverage are outcomes; current-run completion, throughput, and new-pattern intervals are leading indicators.

Graph quality

Not measured yet. The producer must publish exact graph matches, weighted defects, and valid-graph coverage with one frozen corpus/parser/Tree++/grammar/engine identity.

Reference resolution

Not measured yet. The coreference producer has not published its fixed-corpus scorecard.

How to read this: “Coloured” means a rule has explained observed corpus text; “Closed” means a rule claims every final action in the sentence needs no more processing; it is a claim, not a proof. Neither percentage is a claim that the grammar is complete.

How much work are the rules doing?

This is activity, not a completion score: it counts how many changes rules make and how many sentences each rule touches.

Changes per sentence
1,317 total changes · 2.634 average · 2.0 middle value
0
1
2
3
4
5
6
7
8
9
10
11
12
14
15
Each bar shows how many sentences received that many changes.
Sentences matched by each rule
RuleSentences matchedChanges made
OP-SEM-POSSESSIVE-OWNER-V2107107
OP-SEM-NEGATION-NOT-SCOPE-V14747
OP-SEM-CARDINAL-QUANTITY-V14444
OP-SEM-COPULA-QUALIFIER-V22222
OP-STR-014-SHORT-V21919
OP-SEM-ORDINAL-NOMINAL-V11111
OP-STR-013-V21010
OP-QUOTE-NESTED-FINITE-COMPLEMENT-V199
OP-SEM-EXISTENTIAL-V299
OP-STR-211-split-V299
OP-SAY-FINITE-REPORTED-CONTENT-V277
OP-SAY-POSTPOSED-DIRECT-QUOTE-V277
OP-SEM-ADJECTIVAL-PASSIVE-STATE-V277
OP-STR-072-V2a77
OP-SEM-PASSIVE-VOICE-V266
OP-STR-017c-V266
OP-STR-043d-V266
OP-STR-045-V266
OP-NOMINAL-PARTICIPIAL-SINGLE-RESTRICTIVE-SINGULAR-V255
OP-RELCL-CONTACT-GAP-V255
OP-STR-072-V2b55
OP-STR-214-V255
OP-WANT-SUBJECT-CONTROL-DESIDERATIVE-V255
OP-SEM-COPULA-INSTANCE-OF-V244
OP-SEM-POLAR-QUESTION-SCOPE-V144
OP-STR-014-V244
OP-STR-014f-V244
OP-STR-015-V244
OP-STR-016c-where-V244
OP-STR-213-V244
OP-STR-215-V244
OP-STR-216c-V244
OP-STR-EPISTEMIC-ATTITUDE-V244
OP-STR-PARTICIPIAL-MEDIAL-CONTROLLED-V244
OP-ACTIVE-INSTRUMENTAL-PURPOSE-V233
OP-COMP-RELATIVE-PASSIVE-SINGLE-V233
OP-FACTIVE-COGNITION-FINITE-CONTENT-V233
OP-HOPE-FINITE-CONTENT-V233
OP-NOMINAL-PARTICIPIAL-SINGLE-RESTRICTIVE-PLURAL-V233
OP-OPEN-COMPLEMENT-IMPLICATIVE-CONTROL-V233
OP-SEM-APPOSITIVE-IDENTITY-DEFINITE-TITLE-V233
OP-SEM-DITRANSITIVE-TRANSFER-DOUBLE-OBJECT-V233
OP-SEM-IS-BY-V233
OP-SEM-UNIVERSAL-PASSIVE-SUBJECT-SELECTOR-V233
OP-SEM-UNIVERSAL-SUBJECT-SELECTOR-V233
OP-STR-118-V2a33
OP-STR-216b-V233
OP-TRY-SUBJECT-CONTROL-ATTEMPT-V233
OP-ADJECTIVAL-DISPOSITION-CONTROL-V222
OP-ASSISTIVE-CLAUSAL-SUBJECT-BARE-XCOMP-V222
OP-CAPTURE-HOW-FINITE-EMBEDDED-QUESTION-V222
OP-HAVE-TO-NECESSARY-ACTION-V222
OP-NEED-TO-NECESSARY-ACTION-V222
OP-NOMINAL-REDUCED-PASSIVE-ACL-RESTRICTIVE-AGENT-SINGULAR-V222
OP-NOMINAL-REDUCED-PASSIVE-ACL-RESTRICTIVE-SINGULAR-V222
OP-SEM-COPULA-LOCATED-IN-V222
OP-SEM-EXISTENTIAL-OBJECT-SELECTOR-V222
OP-SEM-NEGATIVE-SUBJECT-SELECTOR-V222
OP-SEM-PASSIVE-AGENT-V222
OP-SEM-PAST-PARTICIPLE-STATE-EVENT-AMBIGUITY-V222
OP-SEM-UNIVERSAL-OBJECT-SELECTOR-V222
OP-STR-043g-V222
OP-STR-043j-V222
OP-STR-045f-V222
OP-STR-070c-V222
OP-STR-216c-QUOTED-V222
OP-STR-219-V222
OP-ACL-NOMINAL-CONTROL-CONTENT-V211
OP-ADVCL-COPULAR-HERE-PURPOSE-V211
OP-APPEARANCE-PERFECT-RAISING-COORD-V211
OP-APPEARANCE-SEEMING-CCOMP-V211
OP-ARG-KEEP-OBJECT-PREDICATE-RESULT-STATE-V211
OP-ARG-PARALLEL-PREPOSITIONAL-OR-V211
OP-ASPECTUAL-ONSET-BEGIN-XCOMP-FINITE-CONJ-V211
OP-ASPECTUAL-ONSET-BEGIN-XCOMP-V211
OP-ASPECTUAL-ONSET-START-ING-XCOMP-V211
OP-ASPECTUAL-ONSET-START-TO-XCOMP-V211
OP-BE-TO-SCHEDULED-PASSIVE-ACTION-V211
OP-BECAUSE-OF-NOMINAL-CAUSE-V211
OP-CAPABILITY-ACTION-V211
OP-CCOMP-PERCEPTION-OBJECT-CONTROL-V211
OP-COMP-CONTROLLERLESS-GERUND-COORDINATION-V111
OP-COMP-FRONT-EVEN-IF-CONCESSIVE-V211
OP-COMP-IMPERATIVE-COORDINATED-PURPOSE-V211
OP-COMP-LOOK-LAST-WHILE-FOR-SEMICOLON-V111
OP-COMP-NO-COMMA-COUNTERFACTUAL-IF-V211
OP-COMP-QUALITY-BUT-WAY-SVOT-V111
OP-COMP-STABLE-COPULAR-PASSIVE-CONJUNCT-V211
OP-COMP-STABLE-COVER-HANDLING-XCOMP-V211
OP-COMP-STABLE-DROPS-ENABLING-XCOMP-V211
OP-COMP-STABLE-FRONTED-LOSING-FACING-V211
OP-COMP-STABLE-GO-SEE-PURPOSE-V211
OP-COMP-STABLE-STEP-UP-TO-JOIN-V211
OP-COMP-STABLE-STRANDED-FINITE-XCOMP-V211
OP-COMP-TRAILING-WHILE-GERUND-CONTROL-V211
OP-COPULAR-NOMINAL-EMBEDDED-QUESTION-V211
OP-COPULAR-NOT-BUT-CORRECTIVE-CONTRAST-V211
OP-DESPITE-NOMINAL-CONCESSION-V211
OP-EMBEDDED-HOW-TO-XCOMP-V211
OP-END-UP-RESULTATIVE-GERUND-V211
OP-EXPLAIN-HOW-TO-XCOMP-QUESTION-V211
OP-FACTIVE-CCOMP-TEMPORAL-WHEN-V211
OP-FIND-A-WAY-TO-CONTROL-V211
OP-GOAL-COPULAR-TO-TARGET-V211
OP-GOING-TO-PROSPECTIVE-ACTION-V211
OP-HAVE-FUN-GERUND-ACTIVITY-V211
OP-HAVE-TO-NECESSARY-ACTION-ROOT-COORD-CHAIN-V211
OP-HOPE-SUBJECT-CONTROL-ATTITUDE-V211
OP-HORTATIVE-LET-US-CONTENT-V211
OP-IMPERATIVE-BARE-LET-PERMITTED-V211
OP-INSTRUMENTAL-MEANS-GERUND-V211
OP-LEARN-SUBJECT-CONTROL-ACTION-V211
OP-NOMINAL-REDUCED-PASSIVE-ACL-ELABORATIVE-SINGULAR-V211
OP-NOMINAL-REDUCED-PASSIVE-ACL-RESTRICTIVE-AGENT-PLURAL-V211
OP-NOMINAL-REDUCED-PASSIVE-ACL-RESTRICTIVE-PLURAL-V211
OP-OBJECT-CONTROL-TRUST-QUESTION-ATTITUDE-V211
OP-ONLY-IF-BUT-NECESSARY-CONDITION-V211
OP-OPTATIVE-WISH-FINITE-PAST-CONTENT-V211
OP-OUGHT-TO-NECESSARY-ACTION-V211
OP-PAST-HABITUAL-USED-TO-ASPECT-V211
OP-PREFERENCE-SUBJECT-CONTROL-V211
OP-RELCL-OVERT-OBJECT-GAP-V211
OP-SAY-EXISTENTIAL-REPORTED-CONTENT-V211
OP-SEM-APPOSITIVE-ELABORATION-INDEFINITE-V211
OP-SEM-APPOSITIVE-IDENTITY-DEFINITE-NAME-V211
OP-SEM-APPOSITIVE-IDENTITY-POSSESSIVE-PERSON-V211
OP-SEM-COPULA-DEICTIC-SETTING-V211
OP-SEM-COPULA-LOCATED-ON-V211
OP-SEM-DITRANSITIVE-TRANSFER-TO-DATIVE-PREP-V211
OP-SEM-IMPERATIVE-COPULA-PROPERTY-V211
OP-SEM-MAJORITY-SUBJECT-SELECTOR-V211
OP-SEM-ORDINAL-OUTCOME-V111
OP-SEMANTIC-RESULT-CONTAINER-V211
OP-SHARED-MODAL-NWAY-COORDINATION-V211
OP-SHOW-FINITE-EVIDENTIAL-CONTENT-V211
OP-STOP-GERUND-CESSATION-PERFECT-V211
OP-STOP-GERUND-CESSATION-V211
OP-STR-013-COMP-EXPLICIT-SUBJECT-CONJ-V211
OP-STR-013c-V211
OP-STR-014-DO-NEG-V211
OP-STR-014-MODAL-CCOMP-ROOT-COORD-V211
OP-STR-014c-V211
OP-STR-014h-V211
OP-STR-014k-V211
OP-STR-015-v2-V211
OP-STR-043-BARE-PARTICIPIAL-SHARED-SUBJECT-V211
OP-STR-043J-COMMALESS-ONCE-V211
OP-STR-046-V211
OP-STR-216b-QUOTED-V211
OP-STR-CAUSAL-SO-ADVCL-V211
OP-STR-FRONTED-PARTICIPIAL-FINITE-CCOMP-V211
OP-STR-PARALLEL-FINITE-SVO-GAPPING-V211
OP-STR-PARATAXIS-V211
OP-STR-PASSIVE-CONJUNCT-OWNED-SUBJECT-V211
OP-STR-REPORTING-COMP-V211
OP-STR-SERIAL-PARENTHETICAL-DASH-COORD-V211
OP-STR-SERIAL-SHARED-SUBJECT-ADVCL-V211
OP-STR-SERIAL-SHARED-SUBJECT-NESTED-3-V211
OP-TEMP-CONTROLLED-AFTER-GERUND-V211
OP-TEMP-CONTROLLED-BEFORE-GERUND-V211
OP-UNSURE-HOW-TO-XCOMP-QUESTION-V211
By source file

Average and total changes in each source file.

FileAverage changes / sentenceMiddle valueTotal changes
avarisclari-ccby-2024:L9815.01515
sivanea-ccby-2024:L3414.01414
avarisclari-ccby-2024:L65812.01212
tell-me-a-story-ccby-2024:L2531011.01111
ai-research:L101210.01010
dig-unsettling-ccby-2026:L3410.01010
dig-unsettling-ccby-2026:L4310.01010
dig-unsettling-ccby-2026:L6110.01010
five-minute-mysteries-pd-2026:L287710.01010
internet:L5710.01010
petr-hudecek-work-ccby-2015:L10110.01010
petr-hudecek-work-ccby-2015:L12710.01010
petr-hudecek-work-ccby-2015:L3310.01010
sivanea-ccby-2024:L13310.01010
formal-finally:L1609.099
internet:L59.099
mot-editorials:L26589.099
pratham-storyweaver-en-ccby-2024:L38789.099
tell-me-a-story-ccby-2024:L8129.099
1001-stories-cc0-2026:L1318.088
Depth proxy; exact internal transformation events are not exported yet.

Last complete run: 2h ago · full run

⚠ STALE — KPI full refresh just now · engine 7c9ba736e11439b34c3f303b13dfa4a4303fc9e8 · coverage snapshot engine fc7989e7ee4dff8cc3e1fd60476dcb548ba02db1; live grammar 4209d2283