tools/rebuild_pages.sh rebuilds the site within 5 minutes of a change; by hand: python3 -B tools/site.py.1RB0LF_0RC1RD_0LD0RA_1RC0RE_1RF---_1LA0LB
Where it stands
Cryptid candidate, parked at the observation
Halting hunt H-2 (740 line 362, Chaotic Bell Class 2), 2026-10-13: own exact accelerated model (work/h2model.py: k=4 macro machine, run-length tape, own prover of limited Diff rules with nested rule use) ran 362 to step 2.1182e62873 (86.1M loops, 80 rules, 6.06M rule applications) and 433 (same table from E) to 10^43681: no halt. The model matches direct simulation at every loop boundary to 2e7, tmsim at ~1e11 (both starts), sligocki qsim in 1,179 deep fingerprint comparisons to 10^254 (362) and 597 to 10^217 (433), its own k=8 variant to 10^600, and its rules on constructed tapes. It finds the upstream halter tm128 (this table from F) halting at step 81,366,596,481,702 (F,1). Dynamics: counter runs consumed 2 words per pass, parity decides the next layout each phase (Collatz-like lottery, ~10^0.073 steps per phase): cryptid-like for any nonhalting proof. Details: notes.md, 'Halting hunt: exact model (2026-10-13)'.
About this picture
The standard picture (bbchallenge.org's Default view): each row is the whole tape at one step, time runs downward, and the head's path draws the pattern.
Overview
In one sentence
No summary yet: no observe.json, and status.json gives no relation.
- Status
- Cryptid candidatehalting-hunt exact model: done, no halt · updated 2026-10-13Parked · Halting hunt H-2 (740 line 362, Chaotic Bell Class 2), 2026-10-13: own exact accelerated model (work/h2model.py: k=4 macro machine, run-length tape, own prover of limited Diff rules with nested rule use) ran 362 to step 2.1182e62873 (86.1M loops, 80 rules, 6.06M rule applications) and 433 (same table from E) to 10^43681: no halt.idle since Oct 13 21:09 · status.json 2026-10-13
- Class
- Chaotic Bell Class 2
- On the list
- line 328 of
BB6_holdouts_663.txt· historical 740 line 362
Your calls on file 1
- next
- Optional: more lottery compute (cheap, ~2.6e4 orders of magnitude per hour at the end, slowing as sqrt(time)); 428 can run on the same model unchanged. No nonhalting work (cryptid-like).
Mechanism
Proof & checks
No proof work yet.
Files
machines/<tm>/: 3 entries · raw notes, status, scripts, images
| name | size | changed |
|---|---|---|
| notes.md ↓ | 10.0 KB | 2026-10-13 21:09 |
| status.json ↓ | 2.0 KB | 2026-10-13 21:08 |
| work/ | 41 items | 2026-10-13 21:07 |
status.json in full
{
"tm": "1RB0LF_0RC1RD_0LD0RA_1RC0RE_1RF---_1LA0LB",
"label": "H-2",
"status": "CRYPTID_CANDIDATE",
"stage": "halting-hunt exact model: done, no halt",
"updated": "2026-10-13",
"cryptid_flag": true,
"reason": "Halting hunt H-2 (740 line 362, Chaotic Bell Class 2), 2026-10-13: own exact accelerated model (work/h2model.py: k=4 macro machine, run-length tape, own prover of limited Diff rules with nested rule use) ran 362 to step 2.1182e62873 (86.1M loops, 80 rules, 6.06M rule applications) and 433 (same table from E) to 10^43681: no halt. The model matches direct simulation at every loop boundary to 2e7, tmsim at ~1e11 (both starts), sligocki qsim in 1,179 deep fingerprint comparisons to 10^254 (362) and 597 to 10^217 (433), its own k=8 variant to 10^600, and its rules on constructed tapes. It finds the upstream halter tm128 (this table from F) halting at step 81,366,596,481,702 (F,1). Dynamics: counter runs consumed 2 words per pass, parity decides the next layout each phase (Collatz-like lottery, ~10^0.073 steps per phase): cryptid-like for any nonhalting proof. Details: notes.md, 'Halting hunt: exact model (2026-10-13)'.",
"holdout_line_740": 362,
"family": "same table as 433 (start E in 362's labels); RWLAccHaltv6.v tm128 is the table from F and halts",
"hunt": {
"date": "2026-10-13",
"result": "no_halt_to_horizon",
"horizon_362": "2.1182e62873 steps (exact final step in work/m362_k4.jsonl end line)",
"horizon_433": ">= 10^43681.49 steps (run stopped at 20:37 for validation slots; last progress line in work/m433_k4.jsonl)",
"engine": "work/h2model.py (k=4); cross-checked with sligocki qsim via work/h2q.py",
"certificate_route_if_halt": "RWLAcc62_BigUint solve_halt' only for halts within run-length reach (~1e14 for tm128); deeper halts need a certified replay of the limited Diff rules (does not exist)"
},
"next": "Optional: more lottery compute (cheap, ~2.6e4 orders of magnitude per hour at the end, slowing as sqrt(time)); 428 can run on the same model unchanged. No nonhalting work (cryptid-like)."
}