2015
Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
Introduces deterministic soft and stochastic hard visual attention for image captioning.
Annotated bibliography
Read the methods
in context.
Original formulations and the research connected to them, with complete references and direct paper links.
31 papers
No matching papers. Try another phrase or reset the filters.
2017
Structured Attention Networks
Uses differentiable graphical-model inference to attend over subsequences and syntactic trees.
2018
Latent Alignment and Variational Attention
Models alignment as a latent variable and learns approximate posterior inference through variational attention.
2018
Learning Neural Trans-Dimensional Random Field Language Models with Noise-Contrastive Estimation
Uses an exponentially tilted reference distribution and noise-contrastive estimation for neural random-field language models.
2018
Variational Attention for Sequence-to-Sequence Models
Treats attention context vectors as Gaussian latent variables in a variational encoder-decoder.
2019
Fast Structured Decoding for Sequence Models
Adds approximate first-order CRF inference and dynamic transitions to non-autoregressive sequence models.
2020
Bayesian Attention Modules
Builds distributions over attention weights from normalized reparameterizable variables and learns them with Bayesian regularization.
2020
Cascaded Text Generation with Markov Transformers
Higher-order CRF cascades for parallel text generation, parameterized by one Markov Transformer.
2020
Residual Energy-Based Models for Text Generation
Combines a fixed autoregressive language model with a learned sequence-level energy, NCE training, conditional generation by importance resampling, and likelihood bounds.
2021
OpenMEVA: A Benchmark for Evaluating Open-ended Story Generation Metrics
A benchmark for testing automatic story metrics, including human correlation, generalization, robustness, and sensitivity to different coherence errors.
2021
Residual Energy-Based Models for Text
Studies statistical discrimination of human and model text and incorporates learned discriminators into globally normalized generation.
2021
Sequence-to-Lattice Models for Fast Translation
Candidate lookup, parallel conditional edge scoring, and exact Viterbi decoding replace a deep translation decoder.
2022
Model Criticism for Long-Form Text Generation
A framework for evaluating high-level text structure by comparing real and generated data in latent spaces defined by explicit generative critics. Experiments study discourse coherence, coreference, and topicality.
2023
Contrast-augmented Diffusion Model with Fine-grained Sequence Alignment for Markup-to-Image Generation
FSA-CDM combines fine-grained cross-modal sequence alignment, contextual modeling, and a contrast-augmented diffusion objective for markup-to-image generation.
2023
Input Perturbation Reduces Exposure Bias in Diffusion Models
Studies accumulated sampling errors in diffusion models and trains with perturbed inputs to simulate inference-time prediction errors.
2023
LiveTune: Dynamic Parameter Tuning for Feedback-Driven Optimization
LiveVariables expose optimization parameters through designated system ports for real-time adjustment without restarting the optimization loop.
2023
Markup-to-Image Diffusion Models with Scheduled Sampling
Learns to render LaTeX, HTML table layouts, LilyPond music notation, and SMILES molecules with conditional diffusion. Scheduled sampling exposes denoisers to their own generated inputs to address compounding errors.
2023
Tree Prompting: Efficient Task Adaptation without Fine-Tuning
Learns a reusable decision tree whose internal decisions are prompted language-model calls, with few-shot, instruction and dynamically generated prompt candidates.
2023
Augmenting interpretable models with large language models during training
Aug-Tree uses language-model-generated expansions to construct interpretable text features for tree splits.
2024
GPTree: Towards Explainable Decision-Making via LLM-powered Decision Trees
Builds decision trees with LLM-assisted splits and incorporates expert feedback to revise decision paths.
2024
What Makes a Good Story and How Can We Measure It? A Comprehensive Survey of Story Evaluation
A survey of story-quality criteria, human evaluation, automatic metrics, and evaluation benchmarks.
2025
ACT: Agentic Classification Tree
Learns classification trees of natural-language binary questions through impurity-based split selection and TextGrad refinement.
2025
Energy-Based Diffusion Language Models for Text Generation
Adds a full-sequence residual energy to each diffusion denoising distribution, with autoregressive or NCE parameterizations and parallel importance sampling.
2025
Interactive Training: Feedback-Driven Neural Network Optimization
An open-source framework in which humans and AI agents intervene in ongoing neural-network training through a control server, changing optimizer settings, training data, and model checkpoints.
2026
AI Training Manager: Bounded Closed-Loop Control of Adaptive Training Recipes
An asynchronous LLM supervisor observes training telemetry and proposes constrained changes to a running training recipe, with experiments on deliberately induced supervised-learning and reinforcement-learning failure modes.
2026
Breaking the Factorization Barrier in Diffusion Language Models
CoDD composes factorized diffusion predictions with a tractable probabilistic-circuit prior to model joint token dependencies.
2026
DA-DLM: Explicitly Modeling Token Dependencies in Diffusion Language Models
Adapts directed-acyclic dependency modeling to iterative diffusion using position-oriented candidate groups and fixed-token anchors.
2026
DiffuSpec: Unlocking Diffusion Language Models for Speculative Decoding
Uses diffusion drafting, Causal-consistency Path Search over token lattices, and adaptive draft lengths.
2026
Interactive Training 2: Auditable Control Plane for Live Model Training
A reusable protocol with typed settings, structured actions, application-selected control points, and an ordered journal shared by humans, scripts, and LLM controllers.
2026
LiLiCorr: Lightweight Likelihood Correlation of Parallel Drafts for Speculative Decoding
Scores adjacent compatibility among top-K parallel draft candidates and selects a draft through a cheap greedy walk.
2026
Unified Energy for Invariant and Independent Decoding in Diffusion Language Models
Combines energy corrections for dependency and invariance in diffusion decoding.