ept.git/ep_run/dp_ep.py, branch xiang Unnamed repository; edit this file 'description' to name the repository. dp_ep smoke PASSED: 2-GPU lockstep clean, ~107% linear scaling (allreduce negligible vs seconds-long EP steps) 2026-07-07T02:15:45+00:00 Yuren Hao yurenh2@illinois.edu 2026-07-07T02:15:45+00:00 6604a012b96ea9b18d95fabf152072c8728df31b val 6.0->5.83/100 steps at effective B48, res nominal, ranks never diverged. Speed package final ledger: pack 1.47x X DP ~linear (X optional AA-v2 1.3x). dip_screen.py: ARPACK screener for the dip-farm trajectories (running on freed 1080s, 2 chains). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014FAPDWQ49M5Ye3NpTndTpn 
val 6.0->5.83/100 steps at effective B48, res nominal, ranks never diverged.
Speed package final ledger: pack 1.47x X DP ~linear (X optional AA-v2 1.3x).
dip_screen.py: ARPACK screener for the dip-farm trajectories (running on
freed 1080s, 2 chains).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014FAPDWQ49M5Ye3NpTndTpn
Tier-3 gates: Anderson math-yes/impl-no (parked for v2); bf16polish UNSAFE near-edge (eval-only); dp_ep.py ready 2026-07-06T14:42:52+00:00 Yuren Hao yurenh2@illinois.edu 2026-07-06T14:42:52+00:00 40be67d4f5b5a6b46c662c70b759e585e136d70e Anderson: res 25-35x deeper per budget but 5.8x slower (naive history stacks + per-iter safeguard eval) — v2 = ring buffers + periodic safeguard, est +1.3x on the speed tier. bf16+20polish: 1.41x free phase, res parity, BUT z-diff 1.2e-3 — near-marginal operators contract too slowly for a 20-step polish (0.998^20≈0.96), same magnitude as the TF32 kill verdict and the estimator's 50%-sensitivity input. Predicted by our own depth/noise theory. Flags kept with warnings; neither ships for training. dp_ep.py: manual-allreduce EP DP (controller in lockstep, aligned collectives), smoke pending freed 1080s. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_014FAPDWQ49M5Ye3NpTndTpn 
Anderson: res 25-35x deeper per budget but 5.8x slower (naive history stacks
+ per-iter safeguard eval) — v2 = ring buffers + periodic safeguard, est +1.3x
on the speed tier. bf16+20polish: 1.41x free phase, res parity, BUT z-diff
1.2e-3 — near-marginal operators contract too slowly for a 20-step polish
(0.998^20≈0.96), same magnitude as the TF32 kill verdict and the estimator's
50%-sensitivity input. Predicted by our own depth/noise theory. Flags kept
with warnings; neither ships for training. dp_ep.py: manual-allreduce EP DP
(controller in lockstep, aligned collectives), smoke pending freed 1080s.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014FAPDWQ49M5Ye3NpTndTpn