Posters

This paper presents efficient algorithms for transfer learning in latent variable network models, achieving vanishing error under specific conditions, and attaining minimax optimal rates for stochasti…

Supervised learning’s sample complexity is compact: a hypothesis class is learnable if and only if all its finite projections are learnable, simplifying complexity analysis.

This paper introduces transductive active learning, proving its efficiency in minimizing uncertainty and achieving state-of-the-art results in neural network fine-tuning and safe Bayesian optimization…

Generative models can outperform their human trainers: A groundbreaking study shows how autoregressive transformers, trained on chess game data, can achieve higher game ratings than any of the human …

TransAgent empowers vision-language models by collaborating with diverse expert agents, achieving state-of-the-art performance in low-shot visual recognition.

Trajectory Diffusion (T-Diff) significantly improves object goal navigation by learning sequential planning through trajectory diffusion, resulting in more accurate and efficient navigation.

Offline RL with trajectory data achieves statistically efficient learning under linear q*-realizability and concentrability, solving a previously deemed impossible problem.

TrajCLIP: a novel pedestrian trajectory prediction method using contrastive learning and idempotent networks to achieve state-of-the-art performance and enhance generalization across diverse scenarios…

VL-SAM: Training-free open-ended object detection & segmentation using attention maps as prompts, surpassing previous methods on LVIS and CODA datasets.

AdaptiveDiffusion accelerates diffusion model inference by adaptively skipping noise prediction steps, achieving 2-5x speedup without quality loss.

R2ET: training for robust ranking explanations by an effective regularizer.

Researchers reveal how transformers learn word co-occurrence using a novel gradient flow analysis, uncovering a two-phase training process that leads to near-minimum loss and improved model performanc…

VISPA, a novel BNN training framework using Gaussian variational inference and low-rank SDP, achieves state-of-the-art accuracy on various benchmarks.

Train open-vocabulary 3D object detectors using only RGB images and large language models, achieving state-of-the-art performance without expensive LiDAR data.

Train-Attention (TAALM) tackles catastrophic forgetting in LLMs by dynamically weighting tokens during training, boosting learning efficiency and knowledge retention, outperforming existing methods on…

Researchers achieve a balance between accuracy and efficiency in multiclass classification by introducing partially consistent surrogate losses and novel methods.

This paper examines the trade-offs between two popular diagonal Fisher Information Matrix (FIM) estimators in neural networks, deriving variance bounds and highlighting the importance of considering e…

TrAct boosts vision model training by directly optimizing first-layer activations, leading to significant speedups (1.25x-4x) and improved accuracy.

Trace: Automating AI workflow design with LLMs.

Topology-Preserving Reservoirs (TPR) enhances CLIP’s zero-shot learning by using a dual-space alignment and a topology-preserving objective to improve generalization to unseen classes, achieving state…