summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
21 hoursE8 launch order: armB (longest) first, then A/C/D1/D2/D3HEADmainVoid Agent
21 hoursfix(eval): structured-mode greedy decode capped at in-range layout bound ↵Void Agent
(369) — probe-window traces reach 633 tokens and were silently truncated; cap now keys to positional table (1024) so OOD traces decode fully. Regression-verified vs stub model.
43 hoursfix(transformer): increase positional embedding and causal mask buffer to ↵CaptainJack2491
1024 to support worst-case probe traces
43 hoursoptimize E8: precompute causal mask buffer, add eval early exit, and enable ↵CaptainJack2491
compile_model in e8.csv
43 hoursimplement Addendum 8: 6-arm token-space recurrence suite (arms A, B, C, D1, ↵CaptainJack2491
D2, D3), tests, and jobs/e8.csv
44 hoursrefine Addendum 8: expand Arm D into D1 (learned), D2 (frozen), and D3 ↵CaptainJack2491
(dynamic noise)
44 hourspreregister Addendum 8: 4-arm token-space recurrence & decomposition test (E8)CaptainJack2491
45 hoursE7: 20x long-horizon analysis — H-slow rejected (4/4 cells)Void Agent
7 daysE7 relaunch order: GPU cell first, completed tf-wd01 cell last (idempotent skip)Void Agent
7 dayseval: strip torch.compile _orig_mod./module. prefixes when loading ↵Void Agent
checkpoints (salvages compiled-GPU runs)
10 daysrun_sweep: 2 threads for --device cpu jobs (matches Addendum 7 protocol)Void Agent
10 daysE7 jobs: 4M-step long-horizon, mixed-device (rnn->cpu, ↵Void Agent
transformer->cuda+compile)
11 daysfix(train): enforce epoch permutation without replacement, add max cache OOM ↵CaptainJack2491
guard, and handle empty split tensors
11 daysperf(train): VRAM-resident tensor caching, zero-copy GPU batch sampling, and ↵CaptainJack2491
adaptive eval schedule
11 daysfix(eval): correct sieve-rank indexing + test expectationsVoid Agent
- _compute_sieve_rank: find LARGEST k (not smallest) whose signature contains all error preds — fixes P5(4) vs P5(1) bug - _sieve_rank_signature: iterate all small_primes (not just p²≤n) to catch composites like 209=11*19 where 11²>209 - Test expectations updated for [101,200] range (209 outside it) - is_prime task: P1→P3 (no sieve-rank for classification tasks) 48/48 green.
11 daysfix(eval): P5/P6 sieve-rank ladder replaces hardcoded FLAGGED_SIEVE_PREDSVoid Agent
- probe_report: dynamic lo/hi (no more hardcoded [101,200]) - main(): probe range derived from range_end ([range_end+1, range_end+1000]) - P6 = exact (no probe misses), P5(k) = errors match rank-k sieve signature - _sieve_rank_signature: computes composites with all factors > p_k - _compute_sieve_rank: identifies rank from model's error predictions - build_report.py: probe_fig uses sieve rank for coloring - Tests updated for P5/P6 codes (48/48 green)
13 daysE6 Phase 5: complete 16/16 cells (no O1, all P4, no sieve rank)Void Agent
Safety-net run: rsync'd the remaining 8 cells from ichi (wd-1.0 seeds {1,2}, wd-3.0, batch-128), classified the full grid against Addendum 6. Verdict unchanged and now complete: no O1 anywhere, P4 out-of-range in all 16 cells, sieve-rank ladder resolves to 'no sieve' (smallest in-range composite predictions 1001/1003/1010/1079/1099, all small-factor). In-range only: low-wd RNN O-PARTIAL (seed-0), wd-1.0 replication seed-stable O4, batch-128 transformer destabilizes late (train 1.0->0.55, val 0.83->0.45). Tracks e6 summary.csv.
14 daysE6 Phase 5: classify 8/16 finished cells (no O1, all P4, no sieve rank)Void Agent
- Apply Addendum 6 P-ladder on [1001,2000] probe (computed on CPU; eval.py still hardcodes [101,200]) - wd 0.01/0.1/0.3: O-PARTIAL + P4 (val EM 0.77-0.84 in-range); wd 1.0: O4 + P4 - Sieve rank: undefined in all cells (smallest composite pred has small factors; no P5(k)/P6) - 960->967 in all cells is training-split memorization, not sieve evidence - Verdict: data pressure did not change outcome class; walls algorithmic, not data-bound - 4 cells running (wd1.0 seeds 1,2), 4 queued (wd3.0, batch-128) -> 08:30 safety-net cron
2026-08-17README: blog part 1+2 links, part 3 slotVoid Agent
2026-08-17track per-phase summary CSVs (force-added under runs/ ignore)Void Agent
2026-08-17public release prep: full preregistration mirror (addenda 1-6), public ↵Void Agent
README, track per-phase summary CSVs
2026-08-17E6 jobs: 16 cells [2,1000] K=32 cuda AMP (14 primary + 2 batch-128 ↵Void Agent
ablation); Addendum 6 locked 61b002f
2026-08-17perf(cuda): add AMP fp16 autocast with GradScaler and torch.compile ↵CaptainJack2491
reduce-overhead mode
2026-08-17feat(gpu): add CUDA support, chunked eval, bisect prime search, and scaling ↵CaptainJack2491
to 100k (48 tests green)
2026-08-17fix D2: integers-mode probe is N/A by construction (OOV probe inputs), ↵Void Agent
functional regression test; D2 scored (int tokens collapse task: O2, 0-13% vs digits 17-87%); NOTES + report Phase 4 updated; 42 tests
2026-08-16Phase 3 diagnostics results + analysis + report phase 4Void Agent
2026-08-16phase3: task_mode (is_prime), train_frac, lr_decay flags + tests (41 green); ↵Void Agent
Addendum 5 locked; 11-job batch
2026-08-15E2 wd sweep results + analysis + report phase 3Void Agent
2026-08-15E2 wd sweep: 8 jobs (0.01/0.1/0.3/3.0 x both models, seed 0; 1.0 = existing ↵Void Agent
control)
2026-08-15E1 results + analysis + report phase 2Void Agent
2026-08-15speedup: batched eval (46x, sieve-stub verified), run_sweep orchestrator ↵Void Agent
(2-way parallel, idempotent, summaries), E1 jobs; 36 tests
2026-08-14tidy: drop dead --ckpt arg, unused imports/vars; verified report pipeline ↵Void Agent
ad-hoc + 34 suite green
2026-08-14report pipeline: build_report.py (self-contained HTML), eval --runs-dir, ↵Void Agent
plot fix, halting per-input steps
2026-08-14eval: P1 diagnostic prediction-based (flagged set +209), classifier ↵Void Agent
regression suite (10 tests, stub ground truth); 34 tests green
2026-08-14fix codex BLOCKERs: global fixed layout invariance, min_steps off-by-one, ↵Void Agent
integers EOS alias, rerun guard, eval prereg-literal codes + dual-checkpoint honesty; +7 regression tests
2026-08-14rnn: fully-tied cell (read-in+compute+decode), lambda ramp, ACT regression ↵Void Agent
test; design: prior art + review
2026-08-14implement Experiment 1: data pipeline, tied-RNN w/ ACT halting, transformer ↵Void Agent
baseline, train/eval/plot, 16 tests
2026-08-14scaffold: README, LICENSE, design docs (spec + prereg from research repo ↵Void Agent
00c696d)
generated by cgit v1.2.3 (git 2.46.0) at 2026-08-31 17:47:28 +0000