⚠️ EXPERIMENTAL WEIGHTS — NOT A FINAL RELEASE ⚠️

THESE ARE EXPERIMENTAL RESEARCH WEIGHTS. THIS IS NOT A FINISHED OR PRODUCTION MODEL.

Every qwen-3.8-next-40b-exp-* repository is an intermediate checkpoint of an ongoing weight-surgery research line — not a release, not a final version, not a validated product.

  • Not validated for end-user deployment, standalone inference, or any downstream product.
  • No safety, alignment, or quality guarantees. Behaviour can change or regress between checkpoints.
  • Non-standard architecture (qwen4_exp): requires an inference stack that explicitly supports it.
  • Published only for reproducibility of MoE pruning, router-healing, PLE-calibration and distillation experiments.

If you are looking for a model to actually use, use the upstream Qwen release instead.

Qwen-3.8-Next-40B-Exp — Routing Tables & Calibration Artifacts

Artifacts generated during MoE calibration of the qwen-3.8-next-40b-exp-* line: experimental routing-frequency tables, expert salience scores and activation profiling metrics, plus the pruned PLE (Prefill Lookup Embeddings) associative memory table at 2.0M entries.

This repository is not a model. It contains no runnable checkpoint — only tensors and metrics consumed by the surgery and healing scripts.

Contents

File Notes
table-*-of-00004.safetensors PLE n-gram table, 4 shards, 130 tensors, ~10.24 GB
table.safetensors.index.json Weight map for the table

The table is the n-gram associative memory pruned to 2,000,000 base entries (ngram_size: 3, ple_embed_dim: 2560), used by the healed checkpoints of this line.

Experiment line

Repository Role
qwen-3.8-next-40b-exp-clean-bf16 Clean baseline
qwen-3.8-next-40b-exp-clean-healed-bf16 Router-healed, text-only
qwen-3.8-next-40b-exp-moe-pruned-bf16 512→128 expert pruning
qwen-3.8-next-40b-exp-moe-healed-bf16 Pruned + router-healed + PLE 2.0M
qwen-3.8-next-40b-exp-ple-2m-table Routing tables and calibration artifacts
qwen-3.8-next-40b-exp-cartridge-general Delta cartridge testbed
qwen-3.8-next-40b-exp-profiling-metrics Profiling metrics

License

Apache 2.0, inherited from the upstream Qwen base model.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support