Mixtral-ft-MatKG

LoRA adapter for materials knowledge-graph link prediction.

Released for double-blind review of How Large Language Models Learn Materials Science.

Base model	mistralai/Mixtral-8x7B-Instruct-v0.1
PEFT type	LoRA (r=8, alpha=16, dropout=0.05)
Target modules	q_proj, v_proj

Usage

from peft import PeftModel
from transformers import AutoModelForCausalLM
m = AutoModelForCausalLM.from_pretrained("mistralai/Mixtral-8x7B-Instruct-v0.1")
m = PeftModel.from_pretrained(m, "matsci-llm-anon/Mixtral-ft-MatKG")

Citation

@inproceedings{anonymous2026learn,
  title={How Large Language Models Learn Materials Science},
  author={Anonymous},
  booktitle={Under review},
  year={2026}
}

Author and affiliation information is withheld during double-blind review.