Qwen2.5-0.5B-Instruct โ steer-input ONNX (browser activation steering)
The q4f16 ONNX export of Qwen2.5-0.5B-Instruct
with one graph modification: a new input steer of shape [24, 896] (fp32),
added into the residual stream after every transformer layer (cast to fp16 to
match the activations).
steer = zerosโ bit-identical to the unpatched model (verified: max |ฮlogit| = 0.0)steer = coefficient ร vectorโ per-request activation steering, e.g. in onnxruntime-web on WebGPU โ control vectors in a browser tab, no server
Made with patch_onnx.py
for the King of Diamonds in-browser
death game / steering lab.
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐ Ask for provider support
Model tree for Unt1l1f1nd/Qwen2.5-0.5B-Instruct-steer-onnx
Base model
Qwen/Qwen2.5-0.5B Finetuned
Qwen/Qwen2.5-0.5B-Instruct Quantized
onnx-community/Qwen2.5-0.5B-Instruct