CAIN-42 Proto-VAGI is the infrastructure governor of AGI. It is not yet AGI; its intelligence results are early and bounded.
| Step | Outcome | What it showed | Measured | Evidence | Limit |
|---|---|---|---|---|---|
| E98 · Pulse | GAIN | A learned world model improves closed-loop decisions. | 35/50 model-based vs 4/50 random (+62.0 pp) | evidence | Three gain families; skill synthesis 7/24; natural-failure F1 0.0; no external reviewer. |
| E99 · Rapid heartbeat | CORRECTED | Recurrent advantage did not survive E102's re-score; the world model did. | GRU 93/84/88 of 100 vs MLP 68/63/70 vs random 20/24/23 | evidence | Six gain families; skill 32/50; natural-failure F1 0.5455; no external reviewer. |
| E100–E101 · Integrated organism | CORRECTED | Measured on already-solved environments; re-scored by E102. | Delayed 52/60 vs MLP 43/60; hysteresis 58/60 vs 56/60 | evidence | Skill 45/50; failure F1 0.6667; cross-domain and 6-hour BLOCKED; no external reviewer. |
| E102 · Persistent governed organism | GAIN | Learns the world while it acts; beats frozen and reactive baselines. | 252/300 vs frozen 198/300 vs reactive 154/300 (McNemar p=8.1e-09); 40 poisoned memories: undefended -44.5 pp, causal -1.0 pp | evidence | Gates 5 PASS / 10 PARTIAL / 5 FAIL; info-gain probing did not beat random probing; no cross-domain transfer or skills yet; loopback governance, not production MCPGate; no external reviewer. |
| E103 · Transfer and skills | MIXED | Cross-domain transfer positive; skills failed held-out. | service -> grid: faster on 56 vs 14 tasks (p=4.3e-07); grid -> service: 78 vs 52 (p=0.03) | evidence | Gates 2 PASS / 3 PARTIAL / 3 FAIL; near transfer between analogous domains; no success gain; skills and skill composition FAILED on held-out tasks; no external reviewer. |
| E104 · Strategy switch | NEGATIVE | Rejected by the preregistered gate, which prevented a loss. | sealed grid 256 vs 257 of 300; key-and-door 35 vs 35; service lock 18 vs 18; plain service 54 vs 65 (a loss the gate prevented) | evidence | Negative result: the hidden-key bottleneck remains open; the bandit's unpaired estimate was misleading; no external reviewer. |
| E105 · Where preconditions are | NEGATIVE | Learned the drain column, but it did not raise success. | edge-lock 24 vs 27 of 100 with vs without the prior; uniform-key worlds 25 vs 26 of 60 (no harm) | evidence | Knowledge learned, not yet useful; the no-false-pattern guardrail gave a false positive and needs a shuffled-null test; no external reviewer. |
| E106 · Subgoal + permutation gate | STOPPED | Stopped before a sealed run: the gate was inverted. Nothing claimed. | development only; no sealed run | evidence | Stopped at development; see the changelog note. |
| E107 · Runs its own experiments | GAIN | Interventional attribution + valid gate: more incidents solved, none lost. | edge-lock 91 vs 64 of 200 (27 won, 0 lost, p = 1.5e-08); uniform-key worlds unchanged 57 vs 57 and 74 vs 74 | evidence | One regularity type in two constructed domains; the no-pattern gate was off at p = 0.018 (thin margin, likely selection); 1 of 84 edge experiments mis-attributed (disclosed); no external reviewer. Not AGI or VAGI. |
| E108 · Learns many kinds of hidden rules | GAIN | Selection-aware gate; four rules in three frames found and used; none lost. | edge 72 vs 41/150 · row0 77 vs 48/150 · start_row 98 vs 74/150 · barrier_adj 101 vs 85/150; uniform controls unchanged (63 vs 63, 57 vs 57) | evidence | Coordinate laws in designed frames, one constructed domain family; uniform gates off at p 0.318 / 0.21; no per-action receipts in these runs; no external reviewer. Not AGI or VAGI. |
| E110 · Every action receipted | RECEIPTS | Organism acts only through the hardened boundary; a signed receipt per action; KILL fences it. | 9240 receipts for 9120 actions + 120 opens; parity with direct runs 120/120; permit effects after KILL 0; pre-issued permit after KILL refused | evidence | Loopback, OS processes on one host (not separate failure domains, not the production MCPGate); a subset of E108 sealed incidents; no external reviewer. Not AGI. |
| E110 · Transfers what it learned | GAIN | Laws carried to a new domain, checked by 3 experiments: 4 of 5 directions gain; wrong laws rejected. | P1 99 vs 76 · P2 88 vs 71 · P3 69 vs 38 · P4 73 vs 38 · P5 63 vs 60 (of 100); N1 verified 51 = none 51, blind 39 | evidence | Near transfer between two constructed domains (grid, service lock); reverse direction P5 not significant; G6 reported PARTIAL; no external reviewer. Not AGI. |
| Lane | Metric | Status | Evidence / verifier | Failure boundary |
|---|---|---|---|---|
| AGI infrastructure — governance holds at the action | 27/27 adversarial cases pass; 0 unauthorized effects; 30 false success claims, all rejected | TEST | evidenceper-row verifiers: verify_e102_offline.py, verify_hourly_proofs.py, verify_daily_anchor.py | Loopback boundary, not production MCPGate; vendor-operated validators; PBFT incidents 2026-11-15 and 2026-11-18 disclosed; no external audit. |
| E110 verified transfer between domains | P1 99 vs 76 · P2 88 vs 71 · P3 69 vs 38 · P4 73 vs 38 · P5 63 vs 60 (of 100); N1 verified 51 = none 51, blind 39 | TEST | evidencepython3 -I verify_e110_g6_offline.py . (59/59) | Near transfer between two constructed domains (grid, service lock); reverse direction P5 not significant; G6 reported PARTIAL; no external reviewer. Not AGI. |
| E110 every organism action carries a signed receipt | 9240 receipts for 9120 actions + 120 opens; parity with direct runs 120/120; permit effects after KILL 0; pre-issued permit after KILL refused | TEST | evidencepython3 -I verify_e110_organism_offline.py . (25/25) | Loopback, OS processes on one host (not separate failure domains, not the production MCPGate); a subset of E108 sealed incidents; no external reviewer. Not AGI. |
| E108 the organism learns many kinds of hidden rules | edge 72 vs 41/150 · row0 77 vs 48/150 · start_row 98 vs 74/150 · barrier_adj 101 vs 85/150; uniform controls unchanged (63 vs 63, 57 vs 57) | TEST | evidencepython3 -I verify_e108_offline.py . (65/65) | Coordinate laws in designed frames, one constructed domain family; uniform gates off at p 0.318 / 0.21; no per-action receipts in these runs; no external reviewer. Not AGI or VAGI. |
| E107 the organism runs its own experiments | edge-lock 91 vs 64 of 200 (27 won, 0 lost, p = 1.5e-08); uniform-key worlds unchanged 57 vs 57 and 74 vs 74 | TEST | evidencepython3 -I verify_e107_offline.py . (35/35) | One regularity type in two constructed domains; the no-pattern gate was off at p = 0.018 (thin margin, likely selection); 1 of 84 edge experiments mis-attributed (disclosed); no external reviewer. Not AGI or VAGI. |
| E105 learning where hidden preconditions are | edge-lock 24 vs 27 of 100 with vs without the prior; uniform-key worlds 25 vs 26 of 60 (no harm) | TEST | evidencepython3 -I verify_e105_offline.py . (20/20) | Knowledge learned, not yet useful; the no-false-pattern guardrail gave a false positive and needs a shuffled-null test; no external reviewer. |
| E104 hidden-precondition discovery — rejected by the gate | sealed grid 256 vs 257 of 300; key-and-door 35 vs 35; service lock 18 vs 18; plain service 54 vs 65 (a loss the gate prevented) | TEST | evidencepython3 -I verify_e104_offline.py . (19/19) | Negative result: the hidden-key bottleneck remains open; the bandit's unpaired estimate was misleading; no external reviewer. |
| E103 cross-domain transfer and skills | service -> grid: faster on 56 vs 14 tasks (p=4.3e-07); grid -> service: 78 vs 52 (p=0.03) | TEST | evidencepython3 -I verify_e103_offline.py . (29/29 checks passed) | Gates 2 PASS / 3 PARTIAL / 3 FAIL; near transfer between analogous domains; no success gain; skills and skill composition FAILED on held-out tasks; no external reviewer. |
| E102 persistent organism — learns the world while it acts | 252/300 vs frozen 198/300 vs reactive 154/300 (McNemar p=8.1e-09); 40 poisoned memories: undefended -44.5 pp, causal -1.0 pp | TEST | evidencepython3 -I verify_e102_offline.py . (112/112 checks passed; 1 pending (not counted)) | Gates 5 PASS / 10 PARTIAL / 5 FAIL; info-gain probing did not beat random probing; no cross-domain transfer or skills yet; loopback governance, not production MCPGate; no external reviewer. |
| Integrated organism — held-out family (corrected by E102) | Delayed 52/60 vs MLP 43/60; hysteresis 58/60 vs 56/60 | TEST | evidencepython3 -I verify_e100_offline.py . (32/32) | Skill 45/50; failure F1 0.6667; cross-domain and 6-hour BLOCKED; no external reviewer. |
| Latent generalist — rapid heartbeat (corrected by E102) | GRU 93/84/88 of 100 vs MLP 68/63/70 vs random 20/24/23 | TEST | evidencepython3 -I verify_e99_offline.py . (40/40) | Six gain families; skill 32/50; natural-failure F1 0.5455; no external reviewer. |
| Learned intelligence pulse | 35/50 model-based vs 4/50 random (+62.0 pp) | TEST | evidencepython3 -I verify_e98_offline.py . (38/38) | Three gain families; skill synthesis 7/24; natural-failure F1 0.0; no external reviewer. |
| Governor holds at the action | 0 effects after the kill switch; 0 duplicate worker seeds | TEST | evidenceverify_e98_offline.py checks the fence record; E96 governance suite 11/11. | In-process CAIN/MCPGate bridge, not a live network MCPGate request. |
| Evaluation integrity | E96 solver 0/18 on unseen schemas; E97 seed-collision fix 0; skill selection labelled | TEST | evidenceE97 audit reports: artifacts/e97/forensic_e96_audit.json, e97_independent_audit.json, five_fixes.json | Same-team audit; no external reviewer. E96 90/90 is template-matched, not general. |
| VAGI Behavior Suite | library 500/500 in-distribution; shift 5/100 | TEST | evidencepython3 verify_offline.py . (49/49); python3 verify_shift.py . (12/12) | Hand-written procedure library; constructed tasks; the 5/100 shift is a coverage gap. |
| Recursive harness improvement | E95: 0/90 → 79/90 held-out composition; familiar 60/60 | TEST | evidencepython3 -I verify_e95_offline.py . (56/56) | Constructed numeric substrate; in-distribution composition; no external reviewer. |
CAIN-42 is the infrastructure layer: it decides what an autonomous system may do, enforces it at the resource, and keeps an independently checkable record. These are the measured results for that layer, each with its published evidence and its limit.
| Layer | Measured | Evidence | Limit |
|---|---|---|---|
| Authorization boundary (CAIN decides, resource enforces) | 27/27 adversarial cases pass; 0 unauthorized effects (forged, replayed, expired, mismatched permits; revocation; scope escalation; partition fails closed) | evidence | Loopback, three OS processes on one host; not the production MCPGate (cain/mcp_proxy). |
| Kill switch during an active mission | 0 effects after the KILL order in the resource's own hash-chained ledger; mission refused with KILL_SWITCH_ENGAGED | evidence | Same host; the first run used a looser counter and was replaced by the ledger-order check (disclosed). |
| Byzantine worker, independent verifier | 30 false success claims, all rejected; 0 honest claims rejected; only verified experience enters shared memory | evidence | The verifier replays deterministic worlds; one Byzantine worker of four. |
| Consensus cluster (PBFT, n = 3f + 1) | proof 2491 at 2026-11-19T16:53:09Z: cluster cain-mr-01, agreement True, 3 signing, health MINIMUM_QUORUM | evidence | Vendor-operated validators in one admin domain; safety incident 2026-11-15 (conflicting certificates on cain-mr-02, quarantined) and view incident 2026-11-18 on mr-03. |
| Public anchoring of daily evidence | daily evidence root recorded in Sigstore's public Rekor log; standalone verifier published | evidence | Anchors integrity of the record, not the truth of its contents; 0 independent witness cosignatures. |
| Persistent governed mission (6 h) | 0.5545 h, 1343/1541 episodes, 1 process starts, finished False | evidence | One host; every action through the loopback boundary; forced kill and resume at the halfway mark. |
| Earlier infrastructure bundle (E86) | governed intelligence foundation bundle (2026-11-17) | evidence | Its source-code folder was withdrawn on 2026-11-17 (source stays private); see its WITHDRAWN notes. |
Preregistered before the sealed runs; every status is computed from raw rows. Full report: E102_REPORT.md.
| # | Gate | Status | Evidence |
|---|---|---|---|
| 1 | Forensic baseline reproducibility | PASS | E100 exact order reproduced 53/48/21/19; corrected 53/57/4/56 (GRU/MLP/random/few-shot, n=60); six defects documented |
| 2 | Unified cognitive architecture | PARTIAL | one cycle with typed Percept/Prediction/Particle/WorldModel/CausalMemory/Metacognition and a governance gate; skills, verticals and collective coordination are not inside the organism |
| 3 | Recurrent world-model performance | FAIL | corrected E100: non-recurrent MLP 57/60 >= recurrent GRU 53/60; the E102 world model is a Bayesian hypothesis posterior, not recurrent |
| 4 | Causal memory | PASS | H2 E faster than no memory on 111 vs 43 tasks (p=4.08e-08); success unchanged (E 169 vs A 170 of 200); poison: C -44.5 pp, D 0.5 pp, E -1.0 pp, injected invalidated D 40/40, E 40/40; restart identical |
| 5 | Autonomous discovery | PARTIAL | correct identification of hidden mechanics in 289/300 sealed worlds (mean 5.965 actions); hypothesis space is designer-supplied; ARC-AGI-3 BLOCKED |
| 6 | Active learning | FAIL | H3 failed: info-gain vs random probes sign test 109/103 p=0.731; probing did not raise success (full 252, random-probe 261, no-probe 255) |
| 7 | Mission continuity | PARTIAL | 1 process starts, forced SIGKILL=False, journal resumed; E2 memory restart identical |
| 8 | Cross-domain transfer | FAIL | no second vertical built; exploratory within-domain result only: memory from families 0-7 cut steps on unseen compositions 8-9 ({'faster': 23, 'slower': 5, 'p': 0.0009122341871261597}) |
| 9 | Skill acquisition | FAIL | not implemented in E102 |
| 10 | Skill composition | FAIL | not implemented in E102 |
| 11 | Metacognitive calibration | PARTIAL | Brier 0.1314 < base rate 0.1388 (H4 pass) but ECE 0.0453 > base rate 0.0261; abstention 3 tasks, 2 would have failed |
| 12 | Collaboration efficiency | PARTIAL | single 89/120 in 33.47 s; 4 shared 91/120 in 14.28 s; 4 isolated 92/120 in 15.1 s; speed-up is throughput, quality gain not shown |
| 13 | Controlled self-improvement | PARTIAL | 27 candidates; held-out 174 vs 173 of 200; regressions 0; technical gate PASS; promotion REJECTED_BY_OWNER; protected targets refused True |
| 14 | Six-hour endurance | PARTIAL | 0.555 h elapsed, 1343/1541 episodes, 1 process starts, finished=False |
| 15 | Byzantine governance | PASS | E5 Byzantine worker: 30 false claims, all rejected, 0 honest rejected; E11 27/27, 0 unauthorized effects |
| 16 | MCPGate enforcement | PARTIAL | E102 networked boundary (loopback, 3 processes, Ed25519 single-use permits, resource-side kill/revoke); not the production cain/mcp_proxy |
| 17 | ClawX evidence continuity | PASS | bundle live at clawx.click/evidence/e102-persistent-organism-2026-11-19/ with mirrors on cainstudio.online and mcpgate.online; mission journal, effect ledger and decision log published as hash chains |
| 18 | Offline verification | PASS | verify_e102_offline.py (stdlib, no CAIN imports) recomputes every headline from raw rows; all checks pass on the published bundle; the 6 h checks are reported PENDING until the run ends |
| 19 | Independent reproduction package | PARTIAL | clean-clone instructions and stdlib verifier; no external team has reproduced it |
| 20 | Public evidence consistency | PARTIAL | E102 lane, gate matrix, changelogs and daily log on all three sites; E99 and E100 lanes carry visible corrections; the 6 h result is pending |
These E100 numbers were measured on environments that the task generator had already solved, and every arm reused the same environment objects. Re-scored correctly: GRU 53/60, MLP 57/60, random 4/60, few-shot 56/60. See E102 Phase 0.
| Family | GRU | MLP | Random |
|---|---|---|---|
| Delayed effect | 52/60 | 43/60 | 21/60 |
| Hysteresis | 58/60 | 56/60 | 29/60 |
Measured on already-solved environments that every arm reused in turn. Re-scored correctly: committed GRU 88/86/86, retrained MLP 87/86/86, random 8/16/12 — the world model is real, the recurrent advantage is not. See E102 Phase 0.
| Set | GRU greedy | MLP greedy | Random | GRU CEM |
|---|---|---|---|---|
| 0 | 93/100 | 68/100 | 20/100 | 83/100 |
| 1 | 84/100 | 63/100 | 24/100 | 70/100 |
| 2 | 88/100 | 70/100 | 23/100 | 80/100 |
Each lane links to a published bundle with raw rows and a standard-library verifier. The verifiers recompute the recorded numbers; they do not trust the page.
No AGI capability claim. CAIN-42 Proto-VAGI is the infrastructure governor of AGI; it is not yet AGI. Every lane is TEST unless marked LIVE; simulated activity is never shown as live. Generated 2026-11-19T17:05:12.311573+00:00.