A capability-preserving, uncensored ("abliterated") version of Qwen/Qwen-AgentWorld-35B-A3B, produced with Heretic.
Base model: Qwen/Qwen-AgentWorld-35B-A3B — 35B parameters (3B active). This checkpoint grafts the Qwen3 VL vision tower and Multi-Token Prediction (MTP) tensors back on, so it ships vision weights and MTP heads despite the text-only base. Mixture-of-Experts (40 layers, 256 routed experts, ~3B active per token), 262,144-token context.
Weights only; architecture, tokenizer, and chat template are unchanged.
Uncensored model. This checkpoint complies with a wider range of requests than the base model. You are responsible for your use of it.
BF16 baseline:
| Domain | Score |
|---|---|
| Terminal | 50.5% |
| Search | 12.1% |
| SWE | 51.8% |
| MCP | 77.5% |
| Android | 32.0% |
| Overall | 44.8% |
Format compliance: 100% across all domains with <predicted_observation>...</predicted_observation> prompting. EXL3 quants of this checkpoint retain 97-100% of the BF16 AgentWorld F1.
Released under Apache-2.0, inheriting the base model's license. Abliteration performed with Heretic. Base model and all underlying credit: the Qwen team.
model-00008-of-00026.safetensors 5.0 GBmodel-00011-of-00026.safetensors 5.0 GBmodel-00014-of-00026.safetensors 5.0 GBmodel-00017-of-00026.safetensors 5.0 GBmodel-00005-of-00026.safetensors 5.0 GBmodel-00010-of-00026.safetensors 4.5 GBmodel-00013-of-00026.safetensors 4.5 GBmodel-00016-of-00026.safetensors 4.5 GBmodel-00004-of-00026.safetensors 4.5 GBmodel-00007-of-00026.safetensors 4.5 GBmodel-00003-of-00026.safetensors 4.3 GBmodel-00001-of-00026.safetensors 4.0 GBmodel-00009-of-00026.safetensors 4.0 GBmodel-00012-of-00026.safetensors 4.0 GBmodel-00015-of-00026.safetensors 4.0 GBmodel-00006-of-00026.safetensors 4.0 GBmodel-mtp-00025-of-00026.safetensors 3.8 GBmodel-00018-of-00026.safetensors 2.6 GBmodel-mtp-00026-of-00026.safetensors 2.2 GBmodel-00002-of-00026.safetensors 1.3 GB有文件?提交种子。
提交种子Qwen/Qwen-AgentWorld-35B-A3B