referential-dangling-detector

Referential Dangling Dependency Detector

This is the sentence-pair dependency detector released with Relevant but Incomplete: Referential Dangling as a Paradigm-Level Failure Mode in Hard Prompt Compression.

Paper: Relevant but Incomplete: Referential Dangling as a Paradigm-Level Failure Mode in Hard Prompt Compression

Project page: https://cslikai.cn/Referential-Dangling/

The model scores whether a candidate sentence supplies a necessary dependency for a retained sentence, conditioned on the question. It is used by the repository's automatic context-restoration experiments.

Authors

Zhengpei Hu^1,∗, Kai Li^2,∗, Dapeng Fu³, Xuechao Zou², Yuanhao Tang¹, Yue Li¹, Tengfei Cao¹, and Jianqiang Huang^1,†

• ¹School of Computer Technology and Application, Qinghai University
• ²Tsinghua University
• ³Ant Group Security and Intelligence Laboratory (SIL)
• ^∗Equal contribution; ^†corresponding author

Model details

• Architecture: BERT sequence classifier with two labels
• Base model: google-bert/bert-base-uncased
• Labels: NOT_DEPENDENCY (0), DEPENDENCY (1)
• Maximum training input length: 256 tokens
• Input format: retained sentence [SEP] candidate support [SEP] question

Training data

Training pairs were constructed from the HotpotQA training split. Positive pairs contain a retained sentence and a missing gold-support sentence that share a discriminative entity. Negatives include entity-overlapping hard negatives and unrelated deleted sentences. Splitting is grouped by source example to prevent sentence pairs from the same example appearing in both the training and validation partitions.

See src/build_train_tight.py and src/train_detector.py in the Referential-Dangling repository for the data construction and training code.

Usage

import torch
from transformers import AutoModelForSequenceClassification, AutoTokenizer

model_id = "JusperLee/referential-dangling-detector"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForSequenceClassification.from_pretrained(model_id).eval()

retained = "The film was directed by Jane Smith."
candidate = "Jane Smith is a Canadian filmmaker."
question = "What nationality is the film's director?"
text = f"{retained} [SEP] {candidate} [SEP] {question}"
inputs = tokenizer(text, truncation=True, max_length=256, return_tensors="pt")

with torch.no_grad():
    probability = model(**inputs).logits.softmax(dim=-1)[0, 1].item()

print(probability)

For the paper's restoration pipeline, use BertDependencyDetector from src/beaver2_bert.py.

Intended use and limitations

This checkpoint is intended for research on dependency loss and automatic support restoration in compressed English QA contexts. It is not a general factuality, entailment, or coreference model. Its predictions depend on the candidate-generation procedure and may not transfer reliably to other domains, languages, or substantially different compression settings without evaluation.

Citation

@misc{referentialdangling,
  title={Relevant but Incomplete: Referential Dangling as a Paradigm-Level Failure Mode in Hard Prompt Compression},
  author={Zhengpei Hu and Kai Li and Dapeng Fu and Xuechao Zou and Yuanhao Tang and Yue Li and Tengfei Cao and Jianqiang Huang},
  note={Research code and model release}
}