Update README.md

README.md

@@ -35,13 +35,11 @@ Hengchang Liu, Zhao Yang, Bing Su
 
 ---
 
-## 🧠 Method Overview
+## 📝 Abstract (from the paper)
 
-CAL is a training-free framework that:
+Diffusion language models (DLMs) provide a bidirectional generation framework naturally suited for infilling, yet their performance is constrained by the pre-specified infilling length. In this paper, we reveal that DLMs possess an inherent ability to discover the correct infilling length. We identify two key statistical phenomena in the first-step denoising confidence: a local *Oracle Peak* that emerges near the ground-truth length and a systematic *Length Bias* that often obscures this signal. By leveraging this signal and calibrating the bias, our training-free method **CAL** (**C**alibrated **A**daptive **L**ength) enables DLMs to approximate the optimal length through an efficient search before formal decoding. Empirical evaluations demonstrate that CAL improves Pass@1 by up to 47.7\% over fixed-length baselines and 40.5% over chat-based adaptive methods in code infilling, while boosting BLEU-2 and ROUGE-L by up to 8.5% and 9.9% in text infilling. These results demonstrate that CAL paves the way for robust DLM infilling without requiring any specialized training.
 
-* characterizes the *Oracle Peak* phenomenon in diffusion language models,
-* calibrates model confidence with a length bias function,
-* adaptively discovers near-optimal infilling lengths before decoding.
+## 🧠 Evaluation
 
 The method is evaluated on multiple infilling benchmarks including:
 
@@ -66,6 +64,6 @@ If you find this work useful, please cite:
       eprint={2602.00476},
       archivePrefix={arXiv},
       primaryClass={cs.LG},
-      url={https://arxiv.org/abs/2602.00476}
+      url={https://arxiv.org/abs/2602.00476}, 
 }
 ```