研究不同比特分配方式对空间推理效率的影响，发现关键模块精度决定性能。
Where Bits Matter in World Model Planning: A Paired Mixed-Bit Study for Efficient Spatial Reasoning
- 对比均匀、分层、非对称等比特分配策略，分析其对模型表现的影响
- 4比特时性能高度依赖比特分布，6比特以下接近全精度，3比特则严重退化
- 适合关注模型压缩与高效推理的算法研究员和系统工程师
高效的空间推理需要在严格精度预算下仍保持可靠的的世界模型。我们研究低比特规划行为主要由总比特宽度决定，还是由比特在各模块间的分配方式决定。基于 DINO-WM 在 Wall 规划任务上的实验，我们在两种规划器预算下，对统一、混合、非对称及逐层变体进行了成对的混合比特评估。观察到一致的三阶段模式：8比特和6比特设置与FP16相近，3比特设置性能崩溃，而4比特设置对比特分配敏感。在此过渡区，保留编码器精度比均匀量化更有利，近似大小的非对称变体也显示出相同的编码器侧优势。在后续更严格的22单元复制实验中（每单元回合数减少），混合与均匀INT4的差异变为预算相关，进一步凸显该过渡区的敏感性。这些发现推动了模块感知、预算感知的量化策略作为高效空间推理的更广泛研究方向。代码与运行产物见 https://github.com/suraj-ranganath/DINO-MBQuant。
原文摘要 · Abstract (English)
Efficient spatial reasoning requires world models that remain reliable under tight precision budgets. We study whether low-bit planning behavior is determined mostly by total bitwidth or by where bits are allocated across modules. Using DINO-WM on the Wall planning task, we run a paired-goal mixed-bit evaluation across uniform, mixed, asymmetric, and layerwise variants under two planner budgets. We observe a consistent three-regime pattern: 8-bit and 6-bit settings remain close to FP16, 3-bit settings collapse, and 4-bit settings are allocation-sensitive. In that transition region, preserving encoder precision improves planning relative to uniform quantization, and near-size asymmetric variants show the same encoder-side direction. In a later strict 22-cell replication with smaller per-cell episode count, the mixed-versus-uniform INT4 sign becomes budget-conditioned, which further highlights the sensitivity of this transition regime. These findings motivate module-aware, budget-aware quantization policies as a broader research direction for efficient spatial reasoning. Code and run artifacts are available at https://github.com/suraj-ranganath/DINO-MBQuant.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。