Study: continuous diffusion LMs beat discrete ones at few-step structured reasoning via temporal superposition
rishabh16_ · x · 2026-10-14
A systematic theory-plus-experiments study compares continuous and discrete diffusion LMs on reasoning. Key findings:
- Flow-matching models (FLMs) lead discrete diffusion LMs on few-step structured reasoning
- The theory attributes this to temporal superposition: FLMs can keep multiple candidate answers alive in superposition for later reasoning steps
- The work lays out a framework explaining why the two families diverge on reasoning tasks
More from Research
- Best BIPIA prompt-injection detector catches just 2% of AgentDojo attacks at 1% FP — lulzxdxdxd · 2026-10-14
- Robot cornhole debugging: slow-mo wrist cam and 3D KD-tree fix grip and 0.47cm thigh clearance — ki_ki_ki1 · 2026-10-14
- Chollet: Ideal AI-for-Science Outcome Is Millions of AI-Aided Scientists, Not a Few Rich Labs — fchollet · 2026-10-14
- Feeling a dipole's electric field along an ellipse: 2.08kq at the end, 0.13kq on top — TinfoilTricorn · 2026-10-14
- Alibaba's Representation Guidance: plain pixel-diffusion Transformers beat bigger models — bdsqlsz · 2026-10-14
- Visualizing the Fourier series of a sawtooth wave, with Gibbs overshoot hitting ~9% — TinfoilTricorn · 2026-10-14