arxivJul 31
arXiv:2607.25335v2 Announce Type: replace-cross Abstract: Prompt compression shortens LLM input to reduce inference cost, yet existing methods score token importance through LM forward passes. It remains questionable whether such nuanced, costly token selection is necessary. Compression requires ide
arxivJul 31bullish
arXiv:2606.28027v2 Announce Type: replace-cross Abstract: Neural video codecs have surpassed classical codecs in coding efficiency but remain impractical for deployment due to cross-platform incompatibility and high computational cost. Existing quantization-based solutions fail to produce determinis
arxivJul 14bullish
arXiv:2505.18610v2 Announce Type: replace Abstract: Recently, significant progress has been made in developing reasoning-capable Large Language Models (LLMs) through long Chain-of-Thought (CoT) techniques. However, this long-CoT reasoning process imposes substantial memory overhead due to the large
arxivJul 3bullish
arXiv:2607.01520v1 Announce Type: new Abstract: Transformer inference on long sequences is expensive because softmax attention repeatedly reads from a large KV cache. The prevalent approach to this bottleneck is KV cache compression, which replaces the full cache with a compact summary. Despite its
arxivJul 3
arXiv:2607.01517v1 Announce Type: new Abstract: How far can a language model improve under a strict artifact budget? Parameter Golf posed this question as an open community challenge in which participants trained the best language model, with the complete artifact (training code + compressed weights
arxivJun 26
arXiv:2606.26488v1 Announce Type: new Abstract: Recursive reasoning models can solve complex structured tasks with only a few million parameters by repeatedly updating a latent state. Deploying these models on edge hardware requires significant compression, but unlike conventional sequence models, q
arxivJun 18bullish
arXiv:2606.18304v1 Announce Type: cross Abstract: Mixture-of-Experts (MoE) models scale compute efficiently, yet remain expensive to deploy due to their substantial memory footprint and inference overhead. Prior compression methods mainly operate at the expert level, either removing entire experts o
arxivJun 10bullish
arXiv:2606.10450v1 Announce Type: cross Abstract: DiffC provides a principled way to reuse pre-trained diffusion models for lossy compression, but its encoding and decoding procedures remain slow because they require many discretized forward and reverse steps. We study whether few-step generative mo
arxivJun 2bullish
arXiv:2606.02569v1 Announce Type: cross Abstract: Video is temporally redundant: adjacent frames usually share most objects, background, and layout. Yet existing video multimodal large language models (video MLLMs) usually encode each sampled frame as an independent RGB image, causing visual tokens
arxivApr 21bullish
arXiv:2604.15356v1 Announce Type: cross Abstract: Recent work on KV cache quantization, culminating in TurboQuant, has approached the Shannon entropy limit for per-vector compression of transformer key-value caches. We observe that this limit applies to a strictly weaker problem than the one that ac
arxivApr 8bullish
arXiv:2604.05887v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) have advanced unified reasoning over text, images, and videos, but their inference is hindered by the rapid growth of key-value (KV) caches. Each visual input expands into thousands of tokens, causing caches to