BVSR
Collection
Improving Creative Idea Generation in AI Scientists • 3 items • Updated • 2
How to use Xkev/SR-judger with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-classification", model="Xkev/SR-judger") # pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("Xkev/SR-judger")
model = AutoModelForSequenceClassification.from_pretrained("Xkev/SR-judger", device_map="auto")Paper: Blind Variation, Selective Retention: Improving Creative Idea Generation in AI Scientists
Code: https://github.com/zou-group/BVSR
Data: Xkev/BVSR-data
The selective-retention selector of BVSR: it scores a proposal (task specification + code) so that the lowest-ranked proposals can be discarded before verification.
Qwen/Qwen3.5-9B-Basesr_data/train.jsonl.gz of Xkev/BVSR-datalr=1e-5 (head 1e-3), cosine schedule with warm-up, epochs=1Apache-2.0.
Base model
Qwen/Qwen3.5-9B-Base