- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
This paper proposes AMMAS, a multimodal framework for integrating 3D atomic structures and natural language. It interleaves two types of Transformer layers: Shared Generalist Transformers for cross-modal interaction and Modality Expert Transformers for geometry-specific modeling. This design enables geometry-grounded understanding and language-guided generation of continuous 3D structures. Experiments on three newly introduced small-molecule benchmarks demonstrate its effectiveness over existing baselines.
Critical strengths
- The paper presents a clear motivation and a well-structured framework. The overall methodology is clearly explained and effectively illustrated, making the proposed approach easy to follow.
- The paper provides comprehensive experimental evaluations across three benchmarks, with comparisons against various representative baselines. The ablation studies further support the effectiveness of the proposed architectural design.
Critical weaknesses
- I think the current ablations mainly focus on architectural placement, while more in-depth analyses of the co-evolution mechanism and the contribution of continued pretraining (CPT) are still missing. This is particularly important for the generation task, where AMMAS and the Qwen3 baseline differ in several aspects, making it difficult to isolate the contribution of the proposed method.
- Although the authors introduce three new benchmarks, they mainly focus on a relatively narrow set of capabilities, such as geometric classification and constrained conformation editing. I am not sure how well the proposed framework generalizes to broader atomistic science applications. It would also be helpful to evaluate or align the model with more widely adopted benchmarks to better demonstrate its general capabilities.
- While the paper reports substantial improvements in conformation editing, the advantage over BioMatrix under the stricter [email protected] A metric is only around 1 percentage points. I think this deserves further analysis, as the improvement appears much less significant than the gains highlighted elsewhere in the paper. It would be helpful to clarify where the reported performance gains mainly come from.
- Given that several benchmark labels are derived from explicit geometric rules, I think a non-neural geometry-based baseline would be highly informative. Without such a comparison, it is difficult to assess whether the proposed model demonstrates distinctive multimodal reasoning capabilities or mainly learns to approximate deterministic geometric computations that conventional tools can perform directly.
Questions for authors
- Clarify the specific contribution of the proposed co-evolution mechanism beyond the architectural placement of SGT and MET layers, and explain whether the current ablations sufficiently isolate its effectiveness.
- Discuss the contribution of continued pretraining (CPT) to the overall performance, particularly for the generation task, where AMMAS and the Qwen3 baseline differ in several architectural and training components.
- Clarify the generalizability of the proposed framework beyond the three self-constructed benchmarks, and discuss whether evaluation on existing widely adopted benchmarks is feasible.
- Explain why the improvement over BioMatrix under the stricter [email protected] A metric is relatively small, despite the substantial gains reported in other editing metrics.
- Discuss whether conventional geometry-based methods could achieve competitive performance on tasks whose labels are derived from explicit geometric rules, and clarify what additional capabilities AMMAS provides beyond such methods.