Research #Open models Nov 10, 2026 · Originally published Nov 9

Amazon's HOPE prunes MoE experts by modeling how they work in pairs, ranking 1.58 on average at 50% pruning versus 2.42 for the next-best method, REAP

Most of an MoE model's memory goes to idle experts, and cutting them risks capability. A training-free pruning method that composes with quantization, and where it stops working.

Image: Hugging Face blog (Amazon)
Primary source Hugging Face blog (Amazon) · huggingface.co Read the original ↗

// Key points

  • Take Qwen3.5-122B: 256 experts in each MoE layer, only 8 used per token, and 95% of its roughly 250 GB of parameters sit in expert MLPs.
  • HOPE is a second-order method that accounts for pairwise expert interactions; against 4 first-order baselines on Qwen3.5-122B, Qwen3.5-35B and GLM-4.5-Air, it achieved the best average rank of 2.07 across 5 methods and the top finish in 39% of conditions.
  • Its advantage grows with the pruning rate, up to +6.1% on agentic coding tasks over the next-best method; at 10% to 20% pruning, first-order methods are sometimes competitive.
  • A package runs it in three steps, calibrate, solve and prune, and has been tested on Qwen3, Qwen3-Next, Qwen3.5 and GLM-4.5.

Builder's takeWorth a try if you self-host an MoE open model and are memory-bound: calibrate on real traces from your own product (the paper used SWE-Bench trajectories), prune 40% to 50%, stack quantization on top, and validate on your own eval set rather than public leaderboards.

// Background · from #Open models

Full timeline →
  1. Nov 10 NVIDIA Nemotron 3.5 Super VL arrives in SageMaker JumpStart: 120B total and 12B active parameters, a 1M-token context, and an NVFP4 variant of about 75 GB
  2. Nov 9 Ai2 opens ToyBlocks, a robot learning framework: 85,336 long-horizon demonstrations, about 1,600 hours of data, and up to 24x real-time generation on one GPU
  3. Nov 8 Kimi K3.1 Pro (Highspeed) arrives on Amazon Bedrock: open-weight, natively multimodal, a 1-million-token context, and explicit prompt caching

Every Monday, the week in AI in one email

Every Monday morning: the 6 AI stories that mattered most in the past seven days, with my take, in your inbox.

Your email is only used to send the weekly and is never shared. Every email has a one-click unsubscribe link. See the latest issue →