Variational-Ising-Attention:Tailored Attention Matters for Science

arXiv:2607.23634 · cs.LG, cs.AI, physics.chem-ph · Submitted 2026-07-26 · Read on arXiv

cs.LG, cs.AI, physics.chem-ph

Submitted: 2026-07-26

Updated: 2026-10-05

Comments: 24 pages, ~30 figures

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Sparsity — 1× in this paper · explained in 4 episodes

Sources

• Longformer: The Long-Document Transformer
• Big Bird: Transformers for Longer Sequences
• FlashMemory-DeepSeek-V4: Lightning Index Ultra-Long Context via Lookahead Sparse Attention
• QAMA: Scalable Quantum Annealing Multi-Head Attention Operator for Deep Learning
• Retroformer: Pushing the Limits of Interpretable End-to-end Retrosynthesis Transformer
• Order Matters in Retrosynthesis: Structure-aware Generation via Reaction-Center-Guided Discrete Flow Matching

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks