Posters

A globally Q-linearly converging Gauss-Newton method (AGN) is introduced for overparameterized non-convex low-rank matrix sensing, significantly improving convergence compared to existing gradient des…

Researchers achieve globally convergent variational inference by minimizing the expected forward KL divergence, overcoming the limitations of traditional methods.

Restless multi-armed bandits with global rewards (RMAB-G) are introduced, extending the model to handle non-separable rewards and offering novel index-based and adaptive policies that outperform exist…

AI-powered sequence-to-sequence transformers surpass human and algorithmic abilities in discovering Lyapunov functions for dynamical systems, solving a long-standing open problem in mathematics.

Reward-driven distortions in grid cell patterns are global, not local, preserving path integration while encoding environmental landmarks in spatial navigation.

Large-scale Transformer training’s global convergence is proven using weight decay regularization and a refined mean-field analysis, bridging theory and practice.

GLinSAT: A novel neural network layer efficiently solves general linear constraint satisfaction problems via accelerated gradient descent, enabling differentiable backpropagation and improved GPU perf…

UMOD: a novel multi-objective optimization algorithm efficiently generates uniformly distributed Pareto-optimal solutions by maximizing minimal pairwise distances, providing high-quality representatio…

GL-NeRF accelerates NeRF rendering by using Gauss-Laguerre quadrature, drastically reducing MLP calls without needing additional networks or data structures.

GITA, a novel framework, integrates visual graphs into language models for superior vision-language graph reasoning, outperforming existing LLMs and introducing the first vision-language dataset, GVLQ…

GFT: a novel graph foundation model using transferable computation trees as tokens, improving generalization and reducing negative transfer in graph learning.

GFNSeqEditor, a novel biological sequence editing algorithm, efficiently enhances desired properties while minimizing edits using generative flow networks, surpassing existing methods.

Reward learning from human demonstrations enhances supervised fine-tuning (SFT) for better LLM alignment.

GeoNLF: Geometry-guided Pose-free Neural LiDAR Fields revolutionizes LiDAR point cloud processing by cleverly combining neural and geometric optimization for superior novel view synthesis and multi-vi…

Train factorized neural network layers efficiently with Geometry-aware training in Tucker format (TDLRT)!

RNNs represent naturalistic objects in WM using chronological subspaces, defying traditional slot models; object features are less orthogonalized in RNNs vs. perceptual space.

Geometry Cloak embeds invisible perturbations in images to thwart AI-based 3D reconstruction, forcing the AI to generate identifiable patterns that act as watermarks to assert copyright.

Cross-geometry learning using knowledge distillation significantly improves GNN performance by leveraging both Euclidean and hyperbolic geometric properties of graph data.

Improving LLM alignment, this paper introduces soft preference labels & geometric averaging in Direct Preference Optimization, consistently improving performance on standard benchmarks.

GeoTDM: First diffusion model generating realistic 3D geometric trajectories, capturing complex spatial interactions and temporal correspondence, significantly improving generation quality.