Modelpublic

dementor-research/sft_writingprompts_gpt-oss-20b_as_llama-3.1-8b_seed1

sourceHugging Faceupdated 4mo agoView on Hugging Face

0likes17downloads

Model Card

sftwritingpromptsgpt-oss-20basllama-3.1-8b_seed1

LoRA adapter trained via Tinker as part of the dementor intervention-ladder fingerprint persistence study (AAAI 2026 conference).

—Base model: openai/gpt-oss-20b
—Training stage: SFT (LoRA rank 32, target_modules=all-linear)
—Alias: sft_writingprompts_gpt-oss-20b_as_llama-3.1-8b_seed1

Usage

python

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

base = AutoModelForCausalLM.from_pretrained("openai/gpt-oss-20b")
tok = AutoTokenizer.from_pretrained("openai/gpt-oss-20b")
model = PeftModel.from_pretrained(base, "ethantsliu/sft_writingprompts_gpt-oss-20b_as_llama-3.1-8b_seed1")

Part of the dementor matrix: 4 source models × 3 cross-targets × 3 train datasets × 3 seeds × 2 stages = 216 adapters.