LIANG

29

Papers

376

Total Citations

Papers (29)

SuperGPQA: Scaling LLM Evaluation across 285 Graduate Disciplines

NeurIPS 2025arXiv

MMAR: A Challenging Benchmark for Deep Reasoning in Speech, Audio, Music, and Their Mix

NeurIPS 2025arXiv

Hierarchical Temporal Context Learning for Camera-based Semantic Scene Completion

VisualQuality-R1: Reasoning-Induced Image Quality Assessment via Reinforcement Learning to Rank

NeurIPS 2025arXiv

Robots Pre-train Robots: Manipulation-Centric Robotic Representation from Large-Scale Robot Datasets

CyberHost: A One-stage Diffusion Framework for Audio-driven Talking Body Generation

HumanRefiner: Benchmarking Abnormal Human Generation and Refining with Coarse-to-fine Pose-Reversible Guidance

STAMP: Outlier-Aware Test-Time Adaptation with Stable Memory Replay

RoadPainter: Points Are Ideal Navigators for Topology transformER

Sparse Beats Dense: Rethinking Supervision in Radar-Camera Depth Completion

Exploring the Feature Extraction and Relation Modeling For Light-Weight Transformer Tracking

BountyBench: Dollar Impact of AI Agent Attackers and Defenders on Real-World Cybersecurity Systems

NeurIPS 2025arXiv

MimeQA: Towards Socially-Intelligent Nonverbal Foundation Models

NeurIPS 2025arXiv

T2V-OptJail: Discrete Prompt Optimization for Text-to-Video Jailbreak Attacks

NeurIPS 2025arXiv

SMI-Editor: Edit-based SMILES Language Model with Fragment-level Supervision

Learning with Calibration: Exploring Test-Time Computing of Spatio-Temporal Forecasting

NeurIPS 2025arXiv

Transformers Handle Endogeneity in In-Context Linear Regression

AdaLRS: Loss-Guided Adaptive Learning Rate Search for Efficient Foundation Model Pretraining

NeurIPS 2025arXiv

Detoxifying Large Language Models via Autoregressive Reward Guided Representation Editing

NeurIPS 2025arXiv

REGen: Multimodal Retrieval-Embedded Generation for Long-to-Short Video Editing

NeurIPS 2025arXiv

Stable Minima of ReLU Neural Networks Suffer from the Curse of Dimensionality: The Neural Shattering Phenomenon

NeurIPS 2025arXiv

MesaTask: Towards Task-Driven Tabletop Scene Generation via 3D Spatial Reasoning

NeurIPS 2025arXiv

Blackbox Model Provenance via Palimpsestic Membership Inference

NeurIPS 2025arXiv

LEDiT: Your Length-Extrapolatable Diffusion Transformer without Positional Encoding

NeurIPS 2025arXiv

3DSFLabelling: Boosting 3D Scene Flow Estimation by Pseudo Auto-labelling

PlanU: Large Language Model Reasoning through Planning under Uncertainty

NeurIPS 2025arXiv

DSAS: A Universal Plug-and-Play Framework for Attention Optimization in Multi-Document Question Answering

NeurIPS 2025arXiv

NeuroPath: Neurobiology-Inspired Path Tracking and Reflection for Semantically Coherent Retrieval

NeurIPS 2025arXiv

Simulating Society Requires Simulating Thought

NeurIPS 2025arXiv