# Rein-NPC-4B decision model by jmc0 > Rein-NPC adds a decision readout to Qwen3.5-4B through 64 trained token rows. Read its rule tests, local formats and limits on tool calls. Updated 2026-11-12 (Australia/Adelaide) ## Quick answers ### What is Rein-NPC-4B by jmc0? [Rein-NPC-4B](): A Qwen3.5-4B chat checkpoint with a decision readout trained through 64 added token embeddings. [Rein-NPC-4B facts]() [Source]() ### Who makes Rein-NPC-4B? [Rein-NPC-4B](): jmc0 [Rein-NPC-4B facts]() [Source]() By jmc0 · Hugging Face weight release: 2026-11-08 · OPEN A Qwen3.5-4B chat checkpoint with a decision readout trained through 64 added token embeddings. The original backbone weights stay frozen; a separate copy of the output table keeps the trained input rows out of chat generation. Rule questions read option-letter probabilities, while the same model file also writes dialogue. A decision still needs a prefix pass over the conversation; it is not free work inside an ordinary chat response. ## Facts - Base: Qwen/Qwen3.5-4B - Inputs: Text conversations, personas and rule sheets - Runtime: Transformers; a separate GGUF package uses llama.cpp chat and completion endpoints - Licence: Apache-2.0; licence file included - Access: BF16 weights, decision protocol and GGUF variants ## Claims and sources - **VENDOR-REPORTED**: The maker reports 0.958 accuracy and 0.002 ECE on 28,433 synthetic questions about rules excluded from training, with model-audited labels. A separate 1,352-question author-written set scores 0.924. The card's headline and detailed table disagree on some baseline and end-to-end figures, so no numerical end-to-end improvement is asserted here. Situation-triggered tool calls remain a stated weakness. [source]() ## Links - [Source: Card and weights]() - [Source: Licence]() [Other models by jmc0: Rein r1 family]() ## Related projects No related projects are listed in the registry.