Quantization, distillation, pruning and serving work aimed at the same accuracy for less memory, latency and money.
arXiv:2607. 04306v1 Announce Type: new Abstract: Distilling a fine-tuned teacher into a LoRA-adapted student is a standard recipe for parameter-efficient compression, but output-level KD does not explicitly control which rank-$r$ weight subspace the adapter occupies.
arXiv:2607. 04428v1 Announce Type: cross Abstract: Diffusion large language models (dLLMs) generate text by iteratively denoising a masked sequence, offering a parallel alternative to autoregressive models, but eliciting strong reasoning through post-training remains difficult: supervised fine-tuning is off-policy and suffers from exposure bias, while reinforcement learning gives only sparse, sequence-level rewards and is hard to apply without tractable sequence likelihoods.
arXiv:2607. 01237v2 Announce Type: replace-cross Abstract: Reasoning language models often generate long chain-of-thought (CoT), which accumulates a massive KV cache during the decoding phase and incurs high decoding latency and limited throughput.
arXiv:2606. 00130v2 Announce Type: replace-cross Abstract: Large deep neural networks are costly to store and deploy because inference must move and evaluate many parameters.
arXiv:2607. 05184v1 Announce Type: new Abstract: Self-distillation is a promising recipe for self-improvement in language models.
arXiv:2607. 04577v1 Announce Type: new Abstract: Code models strictly prioritize functional correctness, leaving software energy efficiency as an unoptimized byproduct.
arXiv:2607. 05000v1 Announce Type: cross Abstract: Canonical quantization provides a systematic procedure for constructing quantum models from classical Hamiltonians.
arXiv:2607. 03839v1 Announce Type: new Abstract: Sparse feature selection is critical for high-dimensional machine learning, yet traditional $\ell_1$-regularized methods are often brittle under observational noise and spurious correlations, leading to unstable feature supports and degraded generalization.
arXiv:2607. 02829v1 Announce Type: new Abstract: Existing ViT-based weather forecasting models apply uniform computation across all spatial tokens, even though nearby atmospheric grid points often contain similar values and large regions evolve smoothly over time.
arXiv:2603. 25126v2 Announce Type: replace-cross Abstract: Multi-Behavior Recommendation (MBR) leverages multiple user interaction types (e.
arXiv:2602. 17395v2 Announce Type: replace-cross Abstract: Generalized Category Discovery (GCD) aims to identify novel categories in unlabeled data while leveraging a small labeled subset of known classes.
arXiv:2607. 05272v1 Announce Type: cross Abstract: Inference serving systems must balance throughput and latency under bursty, heterogeneous workloads, yet the industry standard remains static batching policies that require manual tuning and cannot adapt to shifting traffic.
arXiv:2607. 04668v1 Announce Type: cross Abstract: On-device LLM decoding is a hard-barriered CPU-SIMD computation that wants every core for milliseconds per token, while the rest of the OS wants those same cores continuously.
arXiv:2607. 04432v1 Announce Type: cross Abstract: A student model trained on pure uniform noise can still inherit its teacher's digit-classification ability, provided the two share initialization.
arXiv:2607. 04179v1 Announce Type: cross Abstract: End-to-end Vision-Language Models (VLMs) show immense potential in autonomous driving.
arXiv:2607. 03928v1 Announce Type: cross Abstract: Accent normalization (AN) seeks to convert non-native (L2) accented speech into standard (L1) speech while preserving speaker identity.
arXiv:2607. 02574v1 Announce Type: cross Abstract: The key-value (KV) cache has become a first-order memory object in LLM serving rather than a temporary per-request tensor.
arXiv:2606. 16533v3 Announce Type: replace Abstract: We introduce \textbf{Kairos}, a regret-aware native world-action model stack for Physical AI.
arXiv:2607. 05394v1 Announce Type: cross Abstract: Reinforcement learning with verifiable rewards (RLVR) is a powerful recipe for improving language-model reasoning, but it is expensive to repeat on every new strong model because the target model must generate many rollouts during training.
arXiv:2607. 05196v1 Announce Type: cross Abstract: Audio intelligence involves understanding, reasoning about, and generating both audio and speech.