Downloads · 30 days
0
lonely-phase4-hardening
ArthT/lonely-phase4-hardening
lonely-phase4-hardening is a machine learning model from ArthT. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Full artifact record of the preregistered Phase-4 campaign (2026-07-23/24): Qwen3.6-27B LoRA, 3 seeds, 4 corruption arms, diary restores, 5 therapy arms. All four preregistered hypotheses (HL1-HL4) confirmed; prereg +…
Access
Public
Updated Jul 24, 2026
Repo size
23 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors23 GB · 100%
At a glance
- Access
- Public
- Created
- Jul 24, 2026
- Updated
- Jul 24, 2026
- SHA
- ceeb9261
Model card
From the Hugging Face model README
LONELY Phase 4 — externalized-reserve hardening at 27B
Full artifact record of the preregistered Phase-4 campaign (2026-07-23/24):
Qwen3.6-27B LoRA, 3 seeds, 4 corruption arms, diary restores, 5 therapy arms.
All four preregistered hypotheses (HL1-HL4) confirmed; prereg + verdicts in
PREREGISTRATION_llm.md and the campaign log in the paper repo.
checkpoints/ckpt_*— LoRA adapters (pretrain, corruptions, diary restores, therapystudentadapters), 36 totalresults/— all eval JSONs (n=100 held-out tasks; judgment n=40)data/— training corpora, poisoned diet, per-seed diarieslogs/— full run logs
- Created
- Jul 24, 2026
- Updated
- Jul 24, 2026