[ACL '26] UNCODE (a.k.a. PC-Sampler): training-free decoding calibration for masked diffusion LLMs. Code for "Empirical Analysis of Decoding Biases in Masked Diffusion Models"
-
Updated
Oct 24, 2026 - Python
[ACL '26] UNCODE (a.k.a. PC-Sampler): training-free decoding calibration for masked diffusion LLMs. Code for "Empirical Analysis of Decoding Biases in Masked Diffusion Models"
UNCode is an online platform for frequent practice and automatic evaluation of computer programming, Jupyter Notebooks and hardware description language (VHDL/Verilog) assignments. Also provides a pluggable interface with your existing LMS.
Project page: Empirical Analysis of Decoding Biases in Masked Diffusion Models (UNCODE, formerly PC-Sampler; ACL 2026)
Multi-Language Encoder/Decoder Tool This project is a web-based tool that allows users to encode or decode text using Base64 encoding. It includes functionality for switching between Arabic and English languages, and provides options for copying the encoded/decoded result to the clipboard.
To associate your repository with the uncode topic, visit your repo's landing page and select "manage topics."