Add paper link and pipeline tag (#1)

- Add paper link and pipeline tag (4242c338bd594a76155c19aa72bf293678d1302b)


Co-authored-by: Niels Rogge <nielsr@users.noreply.huggingface.co>

README.md

@@ -1,20 +1,23 @@
 ---
-license: mit
 base_model: ML-GSAI/LLaDA2.1-mini
 library_name: peft
+pipeline_tag: text-generation
+license: mit
 tags:
-  - lora
-  - diffusion-language-model
-  - long-context
-  - masked-diffusion
-  - memdlm
+- lora
+- diffusion-language-model
+- long-context
+- masked-diffusion
+- memdlm
 ---
 
 # LLaDA2.1-mini-MemDLM
 
 > **Research artifact.** This adapter is released for academic and research purposes. It was trained on a limited dataset as a proof-of-concept for the MemDLM method and is not intended for production use.
 
-A LoRA adapter for [ML-GSAI/LLaDA2.1-mini](https://huggingface.co/ML-GSAI/LLaDA2.1-mini), trained with [MemDLM](https://github.com/JarvisPei/MemDLM) to enhance long-context understanding in discrete diffusion language models.
+This repository contains a LoRA adapter for [ML-GSAI/LLaDA2.1-mini](https://huggingface.co/ML-GSAI/LLaDA2.1-mini), trained using the method introduced in the paper [MemDLM: Memory-Enhanced DLM Training](https://huggingface.co/papers/2603.22241).
+
+MemDLM (Memory-Enhanced Diffusion Language Model) narrows the train-inference gap in Diffusion Language Models (DLMs) by embedding a simulated denoising process into training via Bi-level Optimization. This mechanism enhances long-context understanding and retrieval performance by offloading memorization pressure to model parameters.
 
 ## Usage
 
@@ -43,6 +46,17 @@ bash examples/llada21/eval_run.sh \
     --adapter_model_name_or_path JarvisPei/LLaDA2.1-mini-MemDLM
 ```
 
+## Citation
+
+```bibtex
+@article{pei2026memdlm,
+    title   = {MemDLM: Memory-Enhanced DLM Training},
+    author  = {Zehua Pei and Hui-Ling Zhen and Weizhe Lin and Sinno Jialin Pan and Yunhe Wang and Mingxuan Yuan and Bei Yu},
+    year    = {2026},
+    journal = {arXiv preprint arXiv:2603.22241},
+}
+```
+
 ## License
 
-MIT
+MIT