Nokia Applied Research has released AnyJev, a project that enables users to extract trustworthy decisions directly from an LLM's internal state in a single forward pass.
Researchers present TokenProbe, a framework that addresses the high token consumption of Chain-of-Thought reasoning by exploiting the non-uniform value of tokens within a sequence. The method uses normalized log probability to distinguish core structural tokens from redundant filler, allowing for selective compression.
Independent researcher Mohamed Menasy has published GlucoEdge, an engineering study detailing an end-to-end on-device machine learning pipeline for five-class, 15-minute glucose trend forecasting from continuous glucose monitor data. The work introduces a compact 1D CNN containing only 2,909 parameters and covers the full workflow from PyTorch to LiteRT conversion.
An independent researcher has published a literature review analyzing whether model compression techniques like quantization, pruning, and distillation actually reduce measured energy consumption or merely lower FLOPs.
Multiverse Computing introduces Quantization-Aware Healing (QAH), a method that distills compressed, quantized large language models directly from their original pre-compression teachers. Applied to a GPT-OSS 120B model compressed to 60B parameters and quantized to MXFP4, the approach produces a model that beats its own full-precision bfloat16 version on 7 of 9 benchmarks.
Researchers propose Reduced Matrix Multiplication (RMM), a training-free, input-adaptive inference method that reduces Transformer matrix products by selecting informative slices along contraction dimensions without modifying model weights. Under a simple retention-ratio control, RMM provides a smooth and predictable accuracy-efficiency trade-off across language models ranging from 1B to 70B parameters.
We use cookies to measure traffic and improve the site. You can accept or decline analytics cookies. Privacy policy