- Rating
- 2
- Confidence
- 2
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
In this paper, the authors address the inefficiency of existing inference-time scaling methods for Large Language Model (LLM) reasoning, where width-based and depth-based adaptive strategies are optimized independently, leading to either premature commitment to consensus hallucinations or erroneous pruning of recoverable reasoning paths. To resolve this limitation, the authors propose Dual-Dimensional Consistency (DDC), a unified framework that couples both axes through a shared path-quality signal computed during generation. Specifically, on the width axis, a confidence-weighted Bayesian sequential decision rule terminates sampling only when the leading answer achieves a high-confidence absolute majority; on the depth axis, a trend-aware stratified pruning mechanism distinguishes persistent degradation from transient uncertainty via spectral analysis of the confidence trajectory, gated per query without tuned constants. The budget freed by early termination is redirected to paths still worth extending. Extensive evaluations across five reasoning benchmarks and five LLMs ranging from 1.7B to 32B parameters demonstrate that DDC reduces token consumption by over 10× while matching or exceeding strong adaptive baselines.
Critical strengths
-
The paper proposes a novel unified framework that couples inter-path sampling width and intra-path reasoning depth through a shared confidence signal, addressing the fundamental limitation of existing adaptive inference methods that optimize these two dimensions independently.
-
The proposed DDC method achieves substantial efficiency gains, reducing token consumption by over 10× across five reasoning benchmarks and five LLMs ranging from 1.7B to 32B parameters, while matching or surpassing strong adaptive baselines and demonstrating a highly favorable accuracy-cost trade-off.
Critical weaknesses
-
DDC heavily relies on token confidence to guide termination and pruning. However, confidence may be poorly calibrated, especially in RL-trained reasoning models, which raises concerns about the robustness of the method.
-
The comparison with RASC is conducted only on Qwen3-4B, and since RASC achieves comparable results on GPQA-diamond, the additional benefit of intra-path pruning is not sufficiently demonstrated on other models.
Questions for authors
-
How sensitive is DDC to confidence miscalibration, particularly in RL-trained reasoning models?
-
Can the authors clarify the advantage of DDC over strong width-only baselines such as RASC under comparable compute budgets?
-
Would the reported speedups hold in optimized serving frameworks such as vLLM or SGLang?
-
Does DDC generalize to other tasks, such as code generation, and larger models (e.g., 70B+)?