Scaling-Aware Adapter for Structure-Grounded LLM Reasoning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866918518418571264 |
|---|---|
| author | Jing, Zihao Zeng, Qiuhao Fang, Ruiyi Li, Yan Yi Sun, Yan Wang, Boyu Hu, Pingzhao |
| author_facet | Jing, Zihao Zeng, Qiuhao Fang, Ruiyi Li, Yan Yi Sun, Yan Wang, Boyu Hu, Pingzhao |
| contents | Large language models (LLMs) are enabling reasoning over 2D and 3D structures, yet existing methods remain modality-specific and typically compress structural inputs through sequence-based tokenization or fixed-length query connectors. Such architectures either omit the geometric grounding requisite for mitigating structural hallucinations, or impose inflexible modality fusion bottlenecks that concurrently over-compress and suboptimally allocate structural tokens, thereby impeding the realization of generalized all-atom reasoning. We introduce Cuttlefish, a unified multimodal LLM that grounds language reasoning in geometric cues while scaling modality tokens with structural complexity. First, Scaling-Aware Patching leverages an instruction-conditioned gating mechanism to generate variable-size patches over structural graphs, adaptively scaling the query token budget with structural complexity to mitigate fixed-length connector bottlenecks. Second, Geometry Grounding Adapter refines these adaptive tokens via cross-attention to modality embeddings and injects the resulting modality tokens into the LLM, exposing explicit geometric cues to reduce structural hallucination. Experiments across interdisciplinary all-atom benchmarks demonstrate that Cuttlefish achieves superior performance in heterogeneous structure-grounded reasoning. Code: github.com/zihao-jing/Cuttlefish. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_02780 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Scaling-Aware Adapter for Structure-Grounded LLM Reasoning Jing, Zihao Zeng, Qiuhao Fang, Ruiyi Li, Yan Yi Sun, Yan Wang, Boyu Hu, Pingzhao Artificial Intelligence Machine Learning Large language models (LLMs) are enabling reasoning over 2D and 3D structures, yet existing methods remain modality-specific and typically compress structural inputs through sequence-based tokenization or fixed-length query connectors. Such architectures either omit the geometric grounding requisite for mitigating structural hallucinations, or impose inflexible modality fusion bottlenecks that concurrently over-compress and suboptimally allocate structural tokens, thereby impeding the realization of generalized all-atom reasoning. We introduce Cuttlefish, a unified multimodal LLM that grounds language reasoning in geometric cues while scaling modality tokens with structural complexity. First, Scaling-Aware Patching leverages an instruction-conditioned gating mechanism to generate variable-size patches over structural graphs, adaptively scaling the query token budget with structural complexity to mitigate fixed-length connector bottlenecks. Second, Geometry Grounding Adapter refines these adaptive tokens via cross-attention to modality embeddings and injects the resulting modality tokens into the LLM, exposing explicit geometric cues to reduce structural hallucination. Experiments across interdisciplinary all-atom benchmarks demonstrate that Cuttlefish achieves superior performance in heterogeneous structure-grounded reasoning. Code: github.com/zihao-jing/Cuttlefish. |
| title | Scaling-Aware Adapter for Structure-Grounded LLM Reasoning |
| topic | Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2602.02780 |