From Dev.to Community
Tech Articles
Curated developer articles, tutorials, and guides — auto-updated hourly
Seth WheelerAug 21, 2026 • 8 min read
Error Feedback, Gradient Compression, and Why Adam Breaks It
Error feedback makes a biased gradient compressor unbiased over time, and under SGD it restored the ...
#llm#measurement#quantization#training
5 1
Seth WheelerAug 22, 2026 • 7 min read
Comparing INT4 and NVFP4 Palettes on Real Gradient Tensors
A palette study on synthetic blocks said evenly-spaced INT4 beats NVIDIA's FP4 grid once a Hadamard ...
#llm#quantization#training#measurement
1 2
Seth Wheeler6d ago • 7 min read
A Better FP4 Gradient Quantizer That Training Couldn't Notice
A per-block scale that cuts FP4 gradient-quantization error on 45 of 45 tensors, 14% against the pub...
#llm#measurement#quantization#training
0 0
Breach ProtocolAug 22, 2026 • 4 min read
Qwen3.8-27B's compression data has a blind spot where its speed-up head lives
The importance matrix used to compress Qwen3.8-27B contains no entries for block 64, the model's mul...
#quantization#openweightmodels#qwen#localinference
0 0
Breach ProtocolAug 25, 2026 • 4 min read
LLMs are less resilient to bit flips than accuracy suggests
A supercomputing-conference study that injected more than 13 million simulated hardware faults into ...
#reliability#hardware#inference#quantization
0 0
Breach Protocol3d ago • 4 min read
llama.cpp merged Qwen's new architecture and a 97-gigabyte lookup table
Support for Qwen3.8-Flash-Next landed in llama.cpp on August 27, adding a sparse-attention graph, vi...
#openweights#localinference#llamacpp#quantization
0 0
Breach ProtocolAug 25, 2026 • 4 min read
A 22-gigabyte local coder matched Opus on a 25-problem slice
TielCoder, a 4-bit re-quantization of the Ornith-1.5 mixture-of-experts model, fits in 22.4 GB and f...
#openweightmodels#quantization#localinference#coding
0 0