Results for "structure prediction"

Shift in model outputs.

Training objective where the model predicts the next token given previous tokens (causal modeling).

Deep learning system for protein structure prediction.

Extension of convolution to graph domains using adjacency structure.

Recovering 3D structure from images.

Predicting protein 3D structure from sequence.

A model that assigns probabilities to sequences of tokens; often trained by next-token prediction.

Probabilistic graphical model for structured prediction.

Monte Carlo method for state estimation.

Low-latency prediction per request.

Learning by minimizing prediction error.

Predicting case success probabilities.

GNN using attention to weight neighbor contributions dynamically.

Graphs containing multiple node or edge types with different semantics.

Structured graph encoding facts as entity–relation–entity triples.

Pixel-wise classification of image regions.

Automatically learning useful internal features (latent variables) that capture salient structure for downstream tasks.

When a model cannot capture underlying structure, performing poorly on both training and test data.

Simplified Boltzmann Machine with bipartite structure.

A function measuring prediction error (and sometimes calibration), guiding gradient-based optimization.

Networks with recurrent connections for sequences; largely supplanted by Transformers for many tasks.

Studying internal mechanisms or input influence on outputs (e.g., saliency maps, SHAP, attention analysis).

Feature attribution method grounded in cooperative game theory for explaining predictions in tabular settings.

Local surrogate explanation method approximating model behavior near a specific input.

Inputs crafted to cause model errors or unsafe behavior, often imperceptible in vision or subtle in text.

Error due to sensitivity to fluctuations in the training dataset.

Systematic error introduced by simplifying assumptions in a learning algorithm.

GNN framework where nodes iteratively exchange and aggregate messages from neighbors.

Generating human-like speech from text.

Computing end-effector position from joint angles.