down_proj
PulseAugur coverage of down_proj — every cluster mentioning down_proj across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
Spectral perturbations reveal structured capabilities in Llama and Qwen models
Researchers have developed a method using spectral smoothing and spiking perturbations to analyze the internal structure of large language models. By altering the singular-value spectra of transformer feed-forward proje…
-
On-device LLM safety analysis reveals sparse critical parameters
Researchers have investigated the safety of on-device language models, specifically Llama-2-7B-Chat, to determine if safety-critical parameters are concentrated in sparse subsets. Their analysis revealed that safety sen…
-
vLLM bug allows incompatible LoRA adapters to load silently
A bug in the vLLM library allows users to load LoRA adapters that are not compatible with the specified model architecture, leading to the model defaulting to its base version without any errors or warnings. The issue s…