Machine Learning

Accelerating science with AI and simulations

MIT Professor Rafael Gómez-Bombarelli discusses the transformative potential of AI in scientific research, emphasizing its role in materi...

Improving AI models’ ability to explain their predictions

New technique makes AI models leaner and faster while they’re still learning

[2510.14949] DialectGen: Benchmarking and Improving Dialect Robustness in Multimodal Generation

Abstract page for arXiv paper 2510.14949: DialectGen: Benchmarking and Improving Dialect Robustness in Multimodal Generation

[2510.17018] CoGate-LSTM: Prototype-Guided Feature-Space Gating for Mitigating Gradient Dilution in Imbalanced Toxic Comment Classification

Abstract page for arXiv paper 2510.17018: CoGate-LSTM: Prototype-Guided Feature-Space Gating for Mitigating Gradient Dilution in Imbalanc...

[2509.22630] StateX: Enhancing RNN Recall via Post-training State Expansion

Abstract page for arXiv paper 2509.22630: StateX: Enhancing RNN Recall via Post-training State Expansion

[2509.17183] LifeAlign: Lifelong Alignment for Large Language Models with Memory-Augmented Focalized Preference Optimization

Abstract page for arXiv paper 2509.17183: LifeAlign: Lifelong Alignment for Large Language Models with Memory-Augmented Focalized Prefere...

[2509.02617] Gaussian process surrogate with physical law-corrected prior for multi-coupled PDEs defined on irregular geometry

Abstract page for arXiv paper 2509.02617: Gaussian process surrogate with physical law-corrected prior for multi-coupled PDEs defined on ...

[2506.18601] BulletGen: Improving 4D Reconstruction with Bullet-Time Generation

Abstract page for arXiv paper 2506.18601: BulletGen: Improving 4D Reconstruction with Bullet-Time Generation

[2506.03863] STAR: Learning Diverse Robot Skill Abstractions through Rotation-Augmented Vector Quantization

Abstract page for arXiv paper 2506.03863: STAR: Learning Diverse Robot Skill Abstractions through Rotation-Augmented Vector Quantization

[2504.14135] Unreal Robotics Lab: A High-Fidelity Robotics Simulator with Advanced Physics and Rendering

Abstract page for arXiv paper 2504.14135: Unreal Robotics Lab: A High-Fidelity Robotics Simulator with Advanced Physics and Rendering

[2502.17421] LongSpec: Long-Context Lossless Speculative Decoding with Efficient Drafting and Verification

Abstract page for arXiv paper 2502.17421: LongSpec: Long-Context Lossless Speculative Decoding with Efficient Drafting and Verification

[2411.05183] Why CNN Features Are not Gaussian: A Statistical Anatomy of Deep Representations

Abstract page for arXiv paper 2411.05183: Why CNN Features Are not Gaussian: A Statistical Anatomy of Deep Representations

[2409.01962] Attentive Dilated Convolution for Automatic Sleep Staging using Force-directed Layout

Abstract page for arXiv paper 2409.01962: Attentive Dilated Convolution for Automatic Sleep Staging using Force-directed Layout

[2407.14971] Sim-CLIP: Unsupervised Siamese Adversarial Fine-Tuning for Robust and Semantically-Rich Vision-Language Models

Abstract page for arXiv paper 2407.14971: Sim-CLIP: Unsupervised Siamese Adversarial Fine-Tuning for Robust and Semantically-Rich Vision-...

[2405.06535] Controllable Image Generation with Composed Parallel Token Prediction

Abstract page for arXiv paper 2405.06535: Controllable Image Generation with Composed Parallel Token Prediction

[2305.02657] On the Eigenvalue Decay Rates of a Class of Neural-Network Related Kernel Functions Defined on General Domains

Abstract page for arXiv paper 2305.02657: On the Eigenvalue Decay Rates of a Class of Neural-Network Related Kernel Functions Defined on ...

[2402.15095] The Umeyama algorithm for matching correlated Gaussian geometric models in the low-dimensional regime

Abstract page for arXiv paper 2402.15095: The Umeyama algorithm for matching correlated Gaussian geometric models in the low-dimensional ...

[2512.24062] Energy-Balanced Hyperspherical Graph Representation Learning via Structural Binding and Entropic Dispersion

Abstract page for arXiv paper 2512.24062: Energy-Balanced Hyperspherical Graph Representation Learning via Structural Binding and Entropi...

[2512.15605] Autoregressive Language Models are Secretly Energy-Based Models: Insights into the Lookahead Capabilities of Next-Token Prediction

Abstract page for arXiv paper 2512.15605: Autoregressive Language Models are Secretly Energy-Based Models: Insights into the Lookahead Ca...

[2512.13592] Image Diffusion Preview with Consistency Solver

Abstract page for arXiv paper 2512.13592: Image Diffusion Preview with Consistency Solver

[2512.07988] HOLE: Homological Observation of Latent Embeddings for Neural Network Interpretability

Abstract page for arXiv paper 2512.07988: HOLE: Homological Observation of Latent Embeddings for Neural Network Interpretability

[2511.01831] Routing-Based Continual Learning for Multimodal Large Language Models

Abstract page for arXiv paper 2511.01831: Routing-Based Continual Learning for Multimodal Large Language Models