Google 团队这篇论文解释 DiffusionGemma 的推理黑箱有多大，发现能用 token 瓶颈把深度压到几乎和 Gemma 4 一样，还发现了扩散模型特有的奇怪推理方式。
论文分析 DiffusionGemma 的推理透明度，将其分解为变量透明度和算法透明度。初始发现 DiffusionGemma 的不透明串行深度是自回归 Gemma 4 的 28.6 倍。但通过可解释的 token 瓶颈映射信息流，可将不透明串行深度降至仅 Gemma 4 的 1.1 倍。算法透明度方面，扩散模型因每步所有 token 可变化而更复杂，研究识别了非时间顺序推理、token 与序列涂抹、中间上下文推理等新现象。可监控性测试表明 DiffusionGemma 与 Gemma 4 水平相当。
How Transparent is DiffusionGemma?
LLM reasoning transparency is a critical affordance for understanding model decisions, mitigating misuse and misalignment, and debugging surprising model behaviors. However, DiffusionGemma performs a larger fraction of its computation in a continuous latent space; does this make its reasoning less transparent? We study this question by decomposing transparency into two components: variable transparency, whether we understand intermediate snapshots of a model's computational state; and algorithmic transparency, whether we can use these snapshots to reconstruct the process by which the model arrived at its outputs. Naively, DiffusionGemma has poor variable transparency: its opaque serial depth, the amount of serial computation that occurs in between interpretable model states, seems at first 28.6X higher than the corresponding autoregressive Gemma 4 model. However, we show that we can map the information flowing between denoising steps through an interpretable token bottleneck with no decrease in downstream performance. Treating these intermediate states as interpretable reduces the opaque serial depth to just 1.1X that of Gemma 4. Algorithmic transparency is harder for diffusion models than for autoregressive models because all token predictions in the canvas can change at every denoising step, giving the model the power to implement complicated distributed algorithms during the denoising process. To begin bridging this gap, we conduct a suite of interpretability case studies, uncovering initial evidence of novel diffusion-specific phenomena such as non-chronological reasoning, token and sequence smearing, and intermediate-context reasoning. Finally, we test monitorability, a key application of transparency that measures whether model outputs are useful for downstream tasks. We find that DiffusionGemma is similarly monitorable to Gemma 4.