It Helps to Take a Second Opinion: Teaching Smaller LLMs to Deliberate Mutually via Selective Rationale Optimisation
Comments Accepted at ICLR 2025
期刊&会议
ICLR
International Conference on Learning Representations · 会议 · Machine Learning
MindSimulator: Exploring Brain Concept Localization via Synthetic FMRI
Comments 23 pages, ICLR 2025
Sharpness-Aware Minimization: General Analysis and Improved Rates
Comments 13th International Conference on Learning Representations (ICLR 2025)
CrystalFramer: Rethinking the Role of Frames for SE(3)-Invariant Crystal Structure Modeling
Comments 12 main pages, 3 main figures, and 4 main tables. Published as a conference paper at ICLR 2025. This version moves some appendices into the main text. For more information, see https://omron-sinicx.github.io/crystalframer/
Elliptic Loss Regularization
Comments ICLR 2025
Adversarial Generative Flow Network for Solving Vehicle Routing Problems
Comments Accepted at ICLR 2025
3D-AffordanceLLM: Harnessing Large Language Models for Open-Vocabulary Affordance Detection in 3D Worlds
Comments ICLR
Unposed Sparse Views Room Layout Reconstruction in the Age of Pretrain Model
Comments Accepted by ICLR 2025. Github page:https://github.com/justacar/Plane-DUSt3R
Towards Scalable Topological Regularizers
Comments 31 pages, ICLR 2025 camera-ready version
B-STaR: Monitoring and Balancing Exploration and Exploitation in Self-Taught Reasoners
Comments Published as a conference paper at ICLR 2025
Weak-to-Strong Generalization Through the Data-Centric Lens
Comments ICLR 2025
Gradient-Free Generation for Hard-Constrained Systems
Comments Accepted as an ICLR 2025 conference paper
CoRNStack: High-Quality Contrastive Data for Better Code Retrieval and Reranking
Comments Published as a conference paper at ICLR 2025. First and second author had equal contribution
Locality Alignment Improves Vision-Language Models
Comments ICLR 2025 Camera-Ready
Improving Semantic Understanding in Speech Language Models via Brain-tuning
Comments Published as a conference paper at ICLR 2025
Controllable Safety Alignment: Inference-Time Adaptation to Diverse Safety Requirements
Comments ICLR 2025 camera ready
Surgical, Cheap, and Flexible: Mitigating False Refusal in Language Models via Single Vector Ablation
Comments ICLR 2025
Accelerating Auto-regressive Text-to-Image Generation with Training-free Speculative Jacobi Decoding
Comments ICLR 2025; Codes: https://github.com/tyshiwo1/Accelerating-T2I-AR-with-SJD/
Meta Flow Matching: Integrating Vector Fields on the Wasserstein Manifold
Comments Accepted to ICLR 2025
Variational Best-of-N Alignment
Comments Accepted at ICLR 2025
Let the Code LLM Edit Itself When You Edit the Code
Comments ICLR 2025 Camera Ready
cryoSPHERE: Single-particle heterogeneous reconstruction from cryo EM
Journal ref International Conference on Learning Representations (ICLR), 2025
Spread Preference Annotation: Direct Preference Judgment for Efficient LLM Alignment
Comments ICLR 2025 Oral Presentation, 22 pages
Stochastic Polyak Step-sizes and Momentum: Convergence Guarantees and Practical Performance
Comments 13th International Conference on Learning Representations (ICLR 2025)
Last-Iterate Convergence Properties of Regret-Matching Algorithms in Games
Comments Accepted to ICLR 2025
GRAIN: Exact Graph Reconstruction from Gradients
Comments Published at The Thirteenth International Conference on Learning Representations (ICLR) 2025
How Low Can You Go? Searching for the Intrinsic Dimensionality of Complex Networks using Metric Node Embeddings
Comments Published as a conference paper at ICLR 2025
Quality Measures for Dynamic Graph Generative Models
Comments To appear as a spotlight presentation at ICLR 2025
InversionGNN: A Dual Path Network for Multi-Property Molecular Optimization
Comments ICLR 2025
FlowDec: A flow-based full-band general audio codec with high perceptual quality
Comments Accepted at ICLR 2025
