- Rating
- 2(+1)
- Confidence
- 3
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper studies multi-image visual reasoning that combines concrete entity images with abstract relational diagrams. The authors introduce VIVIANA, which covers entity recognition, cross-image relational reasoning, and subgraph reasoning. The dataset is constructed through sampling and visualization of multimodal knowledge graphs, with a teacher model generating local-to-global CoT supervision. The authors further propose GKGRPO, which switches between two fixed coefficients for the combined local and global keyword reward depending on whether the final answer is fully correct.
Critical strengths
- The task design connects concrete visual recognition with abstract relational reasoning. VIVIANA organizes entity images, visualized triples, and subgraphs into tasks at different semantic granularities. This provides a platform for studying how local perception supports relational composition.
- The experiments cover multiple training methods and external benchmarks. The authors compare supervised fine-tuning, preference optimization, and several reinforcement learning methods at two model scales, and apply the reward design to different optimization algorithms. Results on external benchmarks show that training gains can transfer beyond knowledge graph images, demonstrating the practical value of the dataset and training approach.
Critical weaknesses
- The dataset’s value as a reliable evaluation and training resource is not sufficiently validated. VIVIANA is a central contribution, but the current presentation does not establish whether its tasks genuinely require cross-image visual reasoning (rather than textual dependency) or whether the synthesized intermediate reasoning is reliable. Complete examples, tests of dependence on textual inputs, and a clear assessment of CoT quality are missing.
- The independent value and robustness of the algorithmic contribution remain unclear. GKGRPO achieves small gains over the main RL baselines, but the paper does not report variability across repeated runs. The current evidence is therefore insufficient to establish a consistent advantage of the proposed gating mechanism over existing methods.
- The contributions of the reward components are not sufficiently isolated. The experiments do not separately evaluate the local and global knowledge rewards. They also lack a fixed-weight control and sensitivity analysis of the gating coefficients, making it difficult to distinguish the benefits of answer-conditioned gating from those of adjusting the knowledge reward weight.
Questions for authors
- Please provide representative examples containing the entity images, visualized triples or subgraphs, question, reference answer, and CoT. Please also clarify whether graph nodes contain entity names, images, or both.
- Some tasks may allow textual shortcuts. In Appendix A.2, the T4.2 prompt already provides the relational path in text, potentially allowing the model to answer using text and prior knowledge without integrating multiple images. Please clarify whether the actual inputs include this textual path and report performance without visual inputs.
- Section 4.2 defines the local and global reference keyword sets but does not explain how they are constructed. Please specify their sources and the construction rules for each task.
- Does the knowledge reward assess only keyword occurrence, without checking whether the resulting statements are correct? Statements with opposite meanings, such as “A is located in B” and “A is not located in B,” may contain the same keywords. Answer-conditioned gating cannot distinguish these reasoning traces when their final answers are identical. Please clarify how this limitation affects the validity of the reward as process supervision.
- Please explain the choice of (\beta_1=0.2) and (\beta_2=0.05) and provide a sensitivity analysis. Does answer-conditioned gating still provide consistent gains over a tuned fixed-weight setting where (\beta_1=\beta_2)? This comparison would help separate the contribution of gating from the effect of adjusting the knowledge reward weight.
Dear reviewers,
The rebuttal period is ending soon. Please read the authors rebuttal and check if they addressed your concerns. Please acknowledge the rebuttal and adjust your scores accordingly.
Thanks.
-AC