summaryrefslogtreecommitdiff
path: root/report_explore
AgeCommit message (Expand)Author
2026-03-26Add Phase 10A.6: gain requires trainable depth-aware aux, not semantic creditYurenHao0426
2026-03-26Add Phase 10A.5: blend gain is implicit regularization, not learned creditYurenHao0426
2026-03-26Add Phase 10A: no prefit threshold — even random Vec blend beats DFA by +1.3%YurenHao0426
2026-03-25Add Phase 9A: checkpointed handoff — blend(Vec+DFA) outperforms pure DFAYurenHao0426
2026-03-25Add Phase 8: schedule timing test — online co-learning is the remaining bot...YurenHao0426
2026-03-25Add Phase 7A: snapshot time sweep shows early snapshots have positive held-ou...YurenHao0426
2026-03-25Add Phase 6.5A: same-batch linesearch REVISES Phase 6A conclusionYurenHao0426
2026-03-24Add Phase 6: snapshot exploitability reveals local update rule is the bottleneckYurenHao0426
2026-03-24Add Phase 5: vector field audit, frozen CIFAR transfer, online pilotYurenHao0426
2026-03-24Add Phase 4 diagnostic dissection: frozen credit recovery, online shallow sca...YurenHao0426
2026-03-24Add CIFAR deltaL test (failed) and pivot design memoYurenHao0426
2026-03-24Add exploration visualization: CIFAR depth scan, boundary ablation, synth vs ...YurenHao0426
2026-03-24Add Phase 3 boundary-condition ablation results and combined memoYurenHao0426
2026-03-23Add Phase 1 synthetic ladder results and memoYurenHao0426