close-out: README table numeric order restored; exp004 gap note updated
Browse files
README.md
CHANGED
|
@@ -82,11 +82,11 @@ resolution) so gaps live in a narrow band β the paired design is load-bearing.
|
|
| 82 |
| exp009_bandroles | role objectives: **directional hit 4/4 but noise-adjacent β frequency reweighting too collinear; needs qualitatively different supervision + generation-side gauges (exp010)** | **shipped (2 seeds + rejudge)** |
|
| 83 |
| exp010_controller | **StepGatedSampler ships**: controller lifts grounding +0.089 over frozen; monotonic lesion ladder; HIGH lesion 14x LP-dominant (coarse-to-fine confirmed in image space); eps-trained HIGH band concentrates (diversity = open training goal) | **shipped (candidate)** |
|
| 84 |
| exp011a_fused_multiband | multiband on REAL fused data: **adapters pay 2-3x more; structural story replicates (3/3 surgical, monolith edge persists); blob targets built 100% after a schema lesson** | **shipped (candidate, s0; s1 running)** |
|
| 85 |
-
| exp015_router_v2 | corrected router test: **address-as-key FALSIFIED** (routing excess 2.5e-06 over constant-key null; match advantage β0.0); open form = TE+AMOE joint training | **shipped (candidate, s0)** |
|
| 86 |
-
| exp014_te_dispatch | router-solidifier v1: **honest negative** β state keys already route by prompt; raw-flattened address kills routing (the known high-D disease geometry); null-design lesson β v2 | **shipped (candidate, s0; s1 running)** |
|
| 87 |
| exp013_blob_flow | **CONDITIONING HYPOTHESIS CONFIRMED**: same blob coupling, ~200x the eps effect on the flow substrate (β5.9% vs +0.03%); blob supervision belongs on flow/v-pred trunks | **shipped (candidate, s0; s1 running)** |
|
|
|
|
|
|
|
| 88 |
|
| 89 |
-
Numbering: a number binds to a preregistered design at assignment and is never reassigned β gaps mean reserved-not-yet-run (
|
| 90 |
|
| 91 |
Code for each experiment ships in its folder with `results.json` from the run
|
| 92 |
ledger. Substrate modules in `substrate/`.
|
|
|
|
| 82 |
| exp009_bandroles | role objectives: **directional hit 4/4 but noise-adjacent β frequency reweighting too collinear; needs qualitatively different supervision + generation-side gauges (exp010)** | **shipped (2 seeds + rejudge)** |
|
| 83 |
| exp010_controller | **StepGatedSampler ships**: controller lifts grounding +0.089 over frozen; monotonic lesion ladder; HIGH lesion 14x LP-dominant (coarse-to-fine confirmed in image space); eps-trained HIGH band concentrates (diversity = open training goal) | **shipped (candidate)** |
|
| 84 |
| exp011a_fused_multiband | multiband on REAL fused data: **adapters pay 2-3x more; structural story replicates (3/3 surgical, monolith edge persists); blob targets built 100% after a schema lesson** | **shipped (candidate, s0; s1 running)** |
|
|
|
|
|
|
|
| 85 |
| exp013_blob_flow | **CONDITIONING HYPOTHESIS CONFIRMED**: same blob coupling, ~200x the eps effect on the flow substrate (β5.9% vs +0.03%); blob supervision belongs on flow/v-pred trunks | **shipped (candidate, s0; s1 running)** |
|
| 86 |
+
| exp014_te_dispatch | router-solidifier v1: **honest negative** β state keys already route by prompt; raw-flattened address kills routing (the known high-D disease geometry); null-design lesson β v2 | **shipped (candidate, s0; s1 running)** |
|
| 87 |
+
| exp015_router_v2 | corrected router test: **address-as-key FALSIFIED** (routing excess 2.5e-06 over constant-key null; match advantage β0.0); open form = TE+AMOE joint training | **shipped (candidate, s0)** |
|
| 88 |
|
| 89 |
+
Numbering: a number binds to a preregistered design at assignment and is never reassigned β gaps mean reserved-not-yet-run (exp005 = the SDXL Tree-4a battery, designed; exp004 closed 2026-07-19). Sub-letters (000b) extend a shipped experiment without disturbing its rows.
|
| 90 |
|
| 91 |
Code for each experiment ships in its folder with `results.json` from the run
|
| 92 |
ledger. Substrate modules in `substrate/`.
|