Electronic Design

Try Alternative Bit Packers For Compression

Compression reduces bandwidth and storage requirements by removing redundancy and irrelevancy. Redundancy occurs when data is sent when it’s not needed. Irrelevancy frequently occurs in audio and ...
IEEE

Manhuff: Adaptive Inverted Huffman Coding for High-Capacity Imperceptible Generative Linguistic Steganography

Abstract: Text steganography inherently grapples with a trade-off between embedding capacity and text quality. Although prior work has explored Huffman coding to mitigate this issue, we observe that ...
IEEE

Implementation and Comparison of Semi-Non-Prefix Huffman Coding in Text Compression

Abstract: The paper implemented a Semi-Non-Prefix (SNP) Huffman coding algorithm for lossless text compression and compared it with traditional Run-Length encoding (RLE), Shannon-Fano, LZ77, and LZ78 ...
GitHub

README.md

A straightforward Huffman coding compressor/decompressor written in C++17. Nothing fancy -- just a clean implementation of the classic algorithm that actually works on real files. I built this mostly ...
marktechpost

An End-to-End Coding Guide to NVIDIA KVPress for Long-Context LLM Inference, KV Cache Compression, and Memory-Efficient Generation

In this tutorial, we take a detailed, practical approach to exploring NVIDIA’s KVPress and understanding how it can make long-context language model inference more efficient. We begin by setting up ...
GitHub

4-bit Adaptive Momentum and Activation Compression for Communication-Efficient Distributed LLM Training

TurboQuant-Adam is an early-stage research prototype exploring whether 4-bit adaptive quantization can reduce communication bandwidth (8x) and activation memory (4x) in distributed LLM training while ...