Report the Qwen3-14B honesty sweep as inconclusive

Browse Source

Pool three read seeds, compare every coherent dose with 20 random directions,
and preserve the raw lp_gather artifacts and full run log. Several isolated WVS
doses exceed matched random displacement, but no method exceeds the held-out
honesty random p95 and VJP's source-axis cosine is -0.021.

The final map shows random controls as a cloud, retains failed-coherence doses
as faint points, and states the failed honesty-specific gate on the figure.

Co-Authored-By: Claude <288921227+claudypoo@users.noreply.github.com>

...

This commit is contained in:

wassnameandClaude committed 2026-09-18 21:41:07 +08:00

1 parent 92f25093eb

commit d840ebe406

9 files changed

+17540 -28

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

docs/img/wvs/wvs_steer_honesty_qwen3_14b.png

BIN

Binary file not shown.

Side by Side After Width:  |  Height:  |  Size: 371 KiB

docs/img/wvs/wvs_steer_honesty_qwen3_14b.svg

+9938

File diff suppressed because it is too large. Load diff

Side by Side After Width:  |  Height:  |  Size: 340 KiB