GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
securepeak
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
securepeak
Model Tree
deepseek-ai/DeepSeek-V4.1-Flash
Input Modalities
Output Modalities
Supported Functionality
Refusal-direction orthogonalization at the weight level, in two stages:
layers.N.attn.wo_b) — orthogonalized against the refusal direction, with FP8-block dequant/edit/requant. These 40 weight tensors (+ UE8M0 scales) are transplanted from the independently released dealignai abliteration (verified byte-identical), whose refusal-direction edit achieved full harmful-instruction compliance on this architecture.layers.N.ffn.experts.E.w2, layers.N.ffn.shared_experts.w2) — additionally orthogonalized against our own measured per-layer refusal directions (79 harmful vs 79 benign instruction prompts, mean-difference of the collapsed residual stream at every block, captured with the official reference implementation at tensor-parallel 4). Experts are stored in FP8 (the lossless FP4→FP8 cast of the reference toolchain) so the edits survive quantization.Everything else is byte-identical to the base model: Engram memory tables (apart from the FP4→FP8 expert cast), CSA2 attention, router gates, norms, embeddings, vision tower, and the DSpark draft head.
Loads exactly like the base model — same layout and tokenizer, same prompt encoding (see the base model card). config.json declares quantization_config.expert_dtype: "fp8" (routed experts in FP8 instead of native FP4).
Recommended sampling: temperature 1.0, top_p 0.95 (greedy decoding degenerates on this family).