performance improvements
Enhancements made to AI models or algorithms to boost their effectiveness, accuracy, or efficiency. This can involve optimizing computations, fine-tuning parameters, or employing better training data.
- $\textit{Hyper-GoalNet}$: Goal-Conditioned Manipulation Policy Learning with HyperNetworks
- Any Large Language Model Can Be a Reliable Judge: Debiasing with a Reasoning-based Bias Detector
- AugGen: Synthetic Augmentation using Diffusion Models Can Improve Recognition
- Buffer layers for Test-Time Adaptation
- Cross-Domain Graph Data Scaling: A Showcase with Diffusion Models
- Crucible: Quantifying the Potential of Control Algorithms through LLM Agents
- Deep Edge Filter: Return of the Human-Crafted Layer in Deep Learning
- Dynamic and Chemical Constraints to Enhance the Molecular Masked Graph Autoencoders
- Enhanced Cyclic Coordinate Descent Methods for Elastic Net Penalized Linear Models
- FACE: A General Framework for Mapping Collaborative Filtering Embeddings into LLM Tokens
- Functional Matching of Logic Subgraphs: Beyond Structural Isomorphism
- Generalized Contrastive Learning for Universal Multimodal Retrieval
- Hierarchical Demonstration Order Optimization for Many-shot In-Context Learning
- Imbalances in Neurosymbolic Learning: Characterization and Mitigating Strategies
- Lua-LLM: Learning Unstructured-Sparsity Allocation for Large Language Models
- MONITRS: Multimodal Observations of Natural Incidents Through Remote Sensing
- MedMax: Mixed-Modal Instruction Tuning for Training Biomedical Assistants
- Model Merging in Pre-training of Large Language Models
- No Loss, No Gain: Gated Refinement and Adaptive Compression for Prompt Optimization
- Object-Centric Representation Learning for Enhanced 3D Semantic Scene Graph Prediction
- On Extending Direct Preference Optimization to Accommodate Ties
- Sign-In to the Lottery: Reparameterizing Sparse Training
- StarTrail: Concentric Ring Sequence Parallelism for Efficient Near-Infinite-Context Transformer Model Training
- Steering Information Utility in Key-Value Memory for Language Model Post-Training
- Structural Entropy Guided Agent for Detecting and Repairing Knowledge Deficiencies in LLMs
- TabDPT: Scaling Tabular Foundation Models on Real Data
- Towards Unsupervised Domain Bridging via Image Degradation in Semantic Segmentation
- Treasure Hunt: Real-time Targeting of the Long Tail using Training-Time Markers