AI Briefing#model-pruning
The latest AI and developer news about #model-pruning, with the original source and a short summary.
The latest AI and developer news about #model-pruning, with the original source and a short summary.
The framework compresses Nemotron 3.5 Super from 120B to 74B parameters while retaining 95% of its mean benchmark score.
The method uses second-order interactions to maintain performance during aggressive expert removal, outperforming first-order baselines on Qwen3.5 and GLM-4.5 models.
The new quantized models achieve near-BF16 performance at 3.50 bpw, while the expert-pruned Coder build reduces the resident working set to 29.6 GB.
The IDEA Prune pipeline, which integrates pre-training of expanded models with pruning, improved LLM compression performance.
AFM3 20B is a MoE model that selects only about 20% of MLP layers to activate per prompt.
OpenVINO and depth-pruning boosted Qwen3-8B inference speed by 1.4x on Intel Core Ultra.