CAIN-42 CAIN Studio
CAIN-42 is an AGI–RSI research organism. Governed general-intelligence and recursive-self-improvement research, with signed evidence for every governed action. AGI: NOT ACHIEVED · NOT ESTABLISHED · RSI: EXPERIMENTAL. RSI Lab · Two Labs · AGI Lab · E135 · 35/48 · E136 · tier-12 · E132 · E127 · E126B · APEX · E137 · AGI+RSI Index · Investor
CAIN-42 · RSI RESEARCH LAB AGI: NOT ESTABLISHED · RSI: EXPERIMENTAL CONSTRUCTED DOMAIN · BOUNDED CLAIMS SIGNED EVIDENCE · OFFLINE VERIFIERS

Recursive Self-Improvement, Measured

AGI: NOT ESTABLISHED · RSI: EXPERIMENTAL

CAIN-42's RSI Research Lab studies one question under governance: can a system write executable improvements into itself, and does the improved system solve held-out tasks the previous one provably could not? Every claim below is a sealed evaluation with a committed answer key, a matched control and an offline verifier that imports no CAIN code.

SISTER LABAGI Research Lab →The wider AGI / AGI–RSI Organism evidence programme. THIS LABRSI Research LabThe self-improvement loop, its ceiling, and its controls. OVERVIEWTwo Labs: AGI + RSI →The canonical side-by-side laboratory page.
Best verified RSI result (E135)
35 / 48
sealed tasks solved; tier-6 6/6; tier-8 5/6; full control suite.
Matched-budget control
22 / 48
same library, same nominal budget, old frontier. The frontier adds +13.
Compute-only control
6 / 48
3.44M evaluations, no learned library: no capability beyond frozen.
Offline verification (E135)
1,551 / 1,551
signed checks recomputed from raw rows by a verifier with no CAIN imports; seed-131 replication repeats the direction.
What is established: in a constructed eight-concept domain, a governed loop that writes its own executable macro library and selects a per-task search frontier raises sealed solves from 26 to 30/48 and tier-6 from 2/6 to 6/6. The gain is not explained by more library entries, more generations, more compute, or a smarter meta-selector.
What is not established: AGI, VAGI, unbounded RSI, human-level capability, external replication or production readiness. Tiers 8, 12 and 16 remain unsolved. The domain is constructed. One team wrote the engine, generator and verifier. There is no third-party audit.

The measured progression

EvolutionWhat changedSealed resultWhat it proved / failed
E129First bounded RSI loop: self-written macro library, sealed tiers, frozen/random/flat/compute controls35/48 on the easier tier-1–8 set; tier-4 1.000Working bounded loop. Saturated at generation 2.
E130Scale curriculum (347 tasks), library (512), meta-strategy, 8 generationsmain 26/48; fixed 28/48; flat 24; compute 17; frozen 6Vocabulary scaling hit a ceiling. The meta-improver hurt. One real rollback.
E131Per-task search portfolio hypothesismain 25; fixed 27; plan 8 (invalid harness arm)Failed intervention. Diagnosis: only 6 of 8 concepts were learned; the top-8 missed one.
E132Concept-complete curriculum + pinned alphabet + portfoliomain 30/48; plan 30; matched fixed 28; fixed 27; compute 6Tier-6 6/6 (was 2/6). Search claim positive. Meta selection neutral. Compute alone is nothing.
E1334+4 meet-in-the-middle for tier 8status SOLVED, but 0/10 correct on every tier-8 taskSpurious fits. The MITM returned reversed halves; 16–48 rows did not identify an 8-concept chain.
E135Corrected half-order + active-identifiability task construction + full controls35/48; tier-6 6/6; tier-8 5/6; flat 22; random 15; compute 6RSI claim POSITIVE_BOUNDED (20/20; seeded replication repeats). One tier-8 task abstains; tiers 12/16 remain 0.
E136Deduplicated 8+4 tier-12 frontier (two 4-chunk compositions, fit/validation split)Tier-12 unit: 10/10 on task 0000; campaign 0/6 (cap/state-set truncation)2,159 4-states → 1,632,736 8-states; ~8.3 min/task. The campaign hit the per-task cap and all six tier-12 tasks abstained; 2/6 also exceed int64. Tier 16 not attempted. Not a tier-12 campaign success.
E138Batched 8+4 and Z3 constraint-solver attempts for tier 12No campaign: batched join exceeded 900 s; Z3 encoding exceeded memoryThe Python/NumPy join and the naive SMT encoding are not viable at 12 concepts. Next: compiled join, table-constrained CP, or learned depth-8 chunks. Not a result.

Why this is a real RSI loop (narrow sense)

Verify it yourself

The bundles are public. Each verifier imports no CAIN code.

Evidence bundles

The tier-12 blocker, measured

What works: the deduplicated 8+4 join solved one real tier-12 task 10/10 (2,159 4-states → 1,632,736 8-states, ~8.3 min). What does not: the campaign returned 0/6 because the per-task cap truncates the join; two of six tasks produce values around 10⁸⁰; the batched NumPy rewrite exceeded 900 s; the Z3 encoding exceeded memory. This is a computational wall, not a missing concept. Tiers 12 and 16 remain unsolved in a campaign.

Honest limits

FIRME AI CORPORATION · CAIN-42 RSI Research Lab · Evidence-first, failures visible. Home · AGI Research Lab · Investors