reaperdoesntknow/DualMind
Text Generation • 2B • Updated • 4.35k
One model, two voices. Explore→Examine→Respond. 6 models + DISC paper. DOI: 10.57967/hf/8184 & 10.57967/hf/8194
Note LogicInference-trained. Explore→Examine→Response cognitive loop.
Note DOI: 10.57967/hf/8184. Four-stage pipeline: TKD + DualMind SFT.
Note Higher quality splits than LogicInference variant. Clean transitions.
Note Quantized Opus variant. Local inference via llama.cpp or LM Studio.
Note Edge deployment. Fits on 6GB GPU or CPU-only. Apache 2.0.
Note Third DualMind model — completes the cognitive architecture.
Note Continuous Thought Dynamics — mathematical backbone of DualMind.
Chat with DualMind and see its reasoning steps
Note Live demo. Explore→Examine→Response cognitive loop on ZeroGPU. Try DualMind without downloading.