Dianbo Liu

A Benchmark for Automatic ML Research Agents Highlighting the Importance of Exploration Breadth

Explore-Execute Chain: Towards an Efficient Structured Reasoning Paradigm

HypoSpace: Evaluating LLM Creativity as Set-Valued Hypothesis Generators under Underdetermination

STORI: A Benchmark and Taxonomy for Stochastic Environments

Token Representation Shrinkage Impairs Creativity of Generative Models

Uncertainty-Based Extensible Codebook for Discrete Federated Learning in Heterogeneous Data Silos

Masked Generative Priors Improve World Models Sequence Modelling Capabilities

Uncertainty-Based Extensible Codebook for Discrete Federated Learning in Heterogeneous Data Silos

A Simple Efficiency Incremental Learning Framework via Vision-Language Model with Multi-Adapters

Evolution guided generative flow networks

Improving Discrete Optimisation Via Decoupled Straight-Through Gumbel-Softmax

Brain-inspired continual pre-trained learner via silent synaptic consolidation

CodeUnlearn: Amortized Zero-Shot Machine Unlearning in Language Models Using Discrete Concept

Gaussian Mixture Vector Quantization with Aggregated Categorical Posterior

Generating GFlowNets as You Wish with Diffusion Process

Small features matter: Robust representation for world models