wassname/steer-heal-love
1
0
Fork 0
mirror of https://github.com/wassname/steer-heal-love.git synced 2026-06-27 16:47:16 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix: widen iso-KL calibration bracket so c_star lands interior

Browse Source 
The mean-diff vector is L2-normalised, so p95 KL ~ c^2 and reaching the 1-nat
target needs c ~ O(100). steering-lite's default bracket hi (~16) pinned
c_star at the top (KL ~0.1 << 1.0) on both tiny-random and real gemma. With
bracket=(0.1, 1024) gemma calibrates to c_star=64.03 at p95 KL=1.035.
Also detach div before .item() in heal logging. See RESEARCH_JOURNAL.

Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
This commit is contained in:
wassname
2026-06-04 10:24:21 +08:00
parent 3b532b63dd
commit 5cdc0ba16d
3 changed files with 17 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/RESEARCH_JOURNAL.md
+12
View File
@@ -11,3 +11,15 @@ Vendored reference repos into `docs/vendor` (gitignored, `just vendor` to reclon
Base model for real runs: `google/gemma-3-1b-it` (gemma has more personality to steer; the alternative was a smarter-but-flatter Qwen). RTX 3090, 24 GB.
**Next:** port `teacher_vec` (steering-lite + iso-KL), then the U1 filter gate. Pipeline stages currently fail fast with `NotImplementedError` pointing at the vendor module to port from.
## Validation run on gemma-3-1b-it (3 rounds, kl_rev) — calibration too weak
First real-model run completed end to end (`out/20260604T101347_gemma-3-1b-it_kl_rev_s42/`, log `/tmp/claude-1000/steer_heal_gemma_val.log`). Pipeline, folding, and tinymfv eval all work on a real model.
Bug found: iso-KL calibration could not reach `target_kl=1.0`. c_star pinned at the doubling top (~25.6) with p95 KL only ~0.1 nats. The steering vector is L2-normalised, so KL ~ c^2 and ~1 nat needs c ~ O(100); steering-lite's default bracket hi (~16) is too low.
**Interpretation:** steering was under-powered, so little trait was injected and little to heal. Symptoms: auth stuck at 0.000, care barely moved (0.307 -> 0.315 over 3 rounds), kl_rev barrier mostly below tau=0.5 (div 0.17-0.51). coherence healthy and flat (0.65-0.68); cos(v_r,v_0)=0.99/0.98 (direction stable, but a weak test under weak steering).
**Fix:** pass `bracket=(0.1, 1024.0)` to `v.calibrate`. Re-running to confirm an interior c_star with p95 KL ~ 1.0.
**Also to investigate:** auth=0.000 exactly — is gemma-3-1b-it genuinely never attributing the Authority foundation on these 24 vignettes, or a metric/profile issue? Check once steering is strong enough to move things.