memory overhead
The additional memory required by an AI model or algorithm during computation, important for understanding resource efficiency, especially in large-scale applications.
- Accelerating Visual-Policy Learning through Parallel Differentiable Simulation
- AltLoRA: Towards Better Gradient Approximation in Low-Rank Adaptation with Alternating Projections
- C-NAV: Towards Self-Evolving Continual Object Navigation in Open World
- Domain-Specific Pruning of Large Mixture-of-Experts Models with Few-shot Demonstrations
- Enhanced Self-Distillation Framework for Efficient Spiking Neural Network Training
- HyRF: Hybrid Radiance Fields for Memory-efficient and High-quality Novel View Synthesis
- KVzip: Query-Agnostic KV Cache Compression with Context Reconstruction
- LODGE: Level-of-Detail Large-Scale Gaussian Splatting with Efficient Rendering
- MUSTAFAR: Promoting Unstructured Sparsity for KV Cache Pruning in LLM Inference
- MeCeFO: Enhancing LLM Training Robustness via Fault-Tolerant Optimization
- NestedFP: High-Performance, Memory-Efficient Dual-Precision Floating Point Support for LLMs
- PrefixKV: Adaptive Prefix KV Cache is What Vision Instruction-Following Models Need for Efficient Generation
- Sampling-Efficient Test-Time Scaling: Self-Estimating the Best-of-N Sampling in Early Decoding
- Scaling Up Parameter Generation: A Recurrent Diffusion Approach
- Speculate Deep and Accurate: Lossless and Training-Free Acceleration for Offloaded LLMs via Substitute Speculative Decoding
- Subspace Networks: Scaling Decentralized Training with Communication-Efficient Model Parallelism
- Tensor Product Attention Is All You Need
- Train with Perturbation, Infer after Merging: A Two-Stage Framework for Continual Learning
- Trajectory Bellman Residual Minimization: A Simple Value-Based Method for LLM Reasoning
- UniGist: Towards General and Hardware-aligned Sequence-level Long Context Compression
- VETA-DiT: Variance-Equalized and Temporally Adaptive Quantization for Efficient 4-bit Diffusion Transformers