SigDiffusions: Score-Based Diffusion Models for Time Series via Log-Signature Embeddings
Comments Published at ICLR 2025
期刊&会议
ICLR
International Conference on Learning Representations · 会议 · Machine Learning
MMFakeBench: A Mixed-Source Multimodal Misinformation Detection Benchmark for LVLMs
Comments Accepted by ICLR 2025, Project page: https://liuxuannan.github.io/MMFakeBench.github.io/
Pursuing Feature Separation based on Neural Collapse for Out-of-Distribution Detection
Comments ICLR 2025
Sports-Traj: A Unified Trajectory Generation Model for Multi-Agent Movement in Sports
Comments Accepted by ICLR 2025. Datasets, code, and model weights are available at: https://github.com/colorfulfuture/UniTraj-pytorch
Not All Language Model Features Are One-Dimensionally Linear
Comments Accepted to ICLR 2025. Code and data at https://github.com/JoshEngels/MultiDimensionalFeatures
Knowledge Localization: Mission Not Accomplished? Enter Query Localization!
Comments ICLR 2025 Spotlight
DECO: Unleashing the Potential of ConvNets for Query-based Detection and Segmentation
Comments ICLR 2025
MolSpectra: Pre-training 3D Molecular Representation with Multi-modal Energy Spectra
Comments Accepted by ICLR 2025
DeepRTL: Bridging Verilog Understanding and Generation with a Unified Representation Model
Comments ICLR 2025 Spotlight
The Hyperfitting Phenomenon: Sharpening and Stabilizing LLMs for Open-Ended Text Generation
Comments Under review at ICLR
Interpreting Language Reward Models via Contrastive Explanations
Comments Accepted at ICLR 2025 conference
Spatial-Mamba: Effective Visual State Space Models via Structure-aware State Fusion
Comments Accepted by ICLR 2025
Associative memory and dead neurons
Comments Reviewed in https://openreview.net/forum?id=mkNVPGpEPm, accepted to ICLR 2025
Beyond Linear Approximations: A Novel Pruning Approach for Attention Matrix
Comments ICLR 2025
Balanced Neural ODEs: nonlinear model order reduction and Koopman operator approximations
Comments conference paper acctepd at ICLR 2025 Singapore
Language Imbalance Driven Rewarding for Multilingual Self-improving
Comments Camera ready version for ICLR 2025
Pareto Low-Rank Adapters: Efficient Multi-Task Learning with Preferences
Comments Accepted at ICLR 2025
Partially Observed Trajectory Inference using Optimal Transport and a Dynamics Prior
Comments ICLR 2025
LancBiO: dynamic Lanczos-aided bilevel optimization via Krylov subspace
Comments This paper is a camera-ready version of ICLR 2025
Distilling Reinforcement Learning Algorithms for In-Context Model-Based Planning
Comments ICLR 2025
(Mis)Fitting: A Survey of Scaling Laws
Comments 41 pages, 3 figure, first two authors contributed equally. ICLR, 2025
Fewer May Be Better: Enhancing Offline Reinforcement Learning with Reduced Dataset
Journal ref Published on ICLR 2025
A Theoretical Perspective: How to Prevent Model Collapse in Self-consuming Training Loops
Comments Accepted at ICLR 2025
Differentially Private Federated Learning With Time-Adaptive Privacy Spending
Comments International Conference on Learning Representations (ICLR), April 2025, Singapore
REFINE: Inversion-Free Backdoor Defense via Model Reprogramming
Comments This paper is accept by ICLR 2025. The first two authors contributed equally to this work. Our code is available at BackdoorBox (https://github.com/THUYimingLi/BackdoorBox) and Github repository (https://github.com/WhitolfChen/REFINE). 28 pages
Unveiling the Magic of Code Reasoning through Hypothesis Decomposition and Amendment
Comments ICLR 2025 Poster;23 pages, 7 figures
Systematic Outliers in Large Language Models
Comments Accepted at ICLR 2025. Project Page: https://github.com/an-yongqi/systematic-outliers
Weighted-Reward Preference Optimization for Implicit Model Fusion
Comments Accepted to ICLR 2025
Multimodality Helps Few-shot 3D Point Cloud Semantic Segmentation
Comments Published at ICLR 2025 (Spotlight)
ELICIT: LLM Augmentation via External In-Context Capability
Comments Published at ICLR 2025
