Token Compression

When the visual encoder and the language decoder of a vision-language model (VLM) run on different compute nodes, the intermediate visual-token embeddings become a communicated payload rather than an internal activation. We call such machine-consumed intermediate tensors AI traffic and ask how far they can be compressed with a standardized, training-free codec. We insert ISO/IEC 15938-17 Neural Network Coding (NNC) round trips on the complete visual interface of a Qwen3-VL-8B-Instruct video question answering pipeline, comprising the main visual-token representation and the DeepStack feature streams, while leaving weights, prompts, and generation untouched, and sweep the quantization parameter (QP) over a wide rate range. Closed-ended Video-MME accuracy remains close to the uncompressed reference up to a 98% reduction of the transmitted BF16 tensor and only then collapses; open-ended MLVU generation shows the same plateau-and-collapse profile under an LLM judge. This robustness is not due to near-lossless reconstruction: the decoded tensor is heavily discretized, carries substantial row-wise relative L2 error, and has a visibly steeper singular-value decay than its source. Downstream reasoning therefore depends on coarse structure and relative geometry rather than exact floating-point values, which argues for rate-task rather than rate-distortion optimization of AI traffic codecs.

Token communications realize the semantic communication principle at the granularity of transformer tokens, providing a promising direction for client--server collaborative inference in resource-constrained edge systems. However, directly transmitting token embeddings presents two practical challenges: substantial communication cost and limited interoperability across model-specific token embedding spaces. To address these challenges, we propose a \emph{token-oriented} semantic communication framework. In this framework, token-level task relevance determines which compressed image latents are transmitted, enabling token-granular transmission without directly transmitting token embeddings. The framework is modular, coordinating three pretrained components---a lightweight client-side vision transformer (ViT), a learned image compression (LIC) model, and a large server-side ViT---without end-to-end training. The key enabler is the one-to-one spatial alignment between ViT patch tokens and the LIC latent vectors, which allows token-level task relevance to directly determine which latent vectors are transmitted. Building on this alignment, token-aligned LIC selectively transmits task-relevant latents, layer-selective attention rollout estimates token relevance from a selected range of attention layers in a single forward pass, and surrogate token substitution adapts the frozen server model by optimizing a single learnable token. Experiments on ImageNet show that the proposed framework achieves a more favorable rate--accuracy trade-off than recent semantic communication schemes, hand-crafted codecs, and task-agnostic LIC models.

Token compression has become a key technique for reducing the inference cost of large foundation models, with approaches such as token pruning and KV-cache reuse widely adopted in vision-language models and recently explored for embodied agents. In embodied agents, tokens not only support perception and semantic understanding but also directly affect latency-sensitive closed-loop robot action prediction. Existing schemes typically guide compression using redundancy or importance cues, such as visual similarity, attention scores, and saliency. However, these cues only indirectly measure the key factor for safe compression: how much a token can change before causing an unacceptable deviation in downstream actions. This receiver-dependent tolerance is closely related to the principle of just noticeable difference (JND). Classical JND characterizes signal tolerance in the human visual system, while machine-oriented JND extends this concept to downstream machine responses. Building on this progression, we introduce Action-JND, which extends JND modeling to embodied perception by defining noticeability through the language-conditioned action response of a vision-language-action (VLA) policy in closed-loop control. A token change is considered admissible only when the induced action deviation remains within a tolerated margin. To realize this concept, we develop a lightweight token-wise JND estimator in deep visual-feature space to predict the maximum tolerable perturbation while preserving policy responses. The resulting action-tolerance score serves as a plug-and-play criterion for VLA compression paradigms, including stale-KV reuse and token pruning, prioritizing action-tolerant tokens for compression. Experiments on the LIBERO benchmark with OpenVLA and OpenVLA-OFT demonstrate that Action-JND consistently improves compression reliability, especially under aggressive compression ratios.

Visual token compression for vision--language models (VLMs) has largely relied on criteria such as attention, redundancy, and uncertainty to maximize average accuracy under a fixed compute budget, implicitly assuming that all errors carry equal cost. However, the consequence of an incorrect prediction on downstream tasks is rarely symmetric: misreading an invoice amount can be far more costly than misclassifying a background color. Motivated by this, we introduce consequence-sensitive visual token compression, which allocates visual computation across requests according to their potential error costs. Our method follows a calibrate-then-allocate procedure, estimating consequence-specific error-budget curves offline and applying the calibrated token budgets online using consequence signals available from question or task information. On a controlled within-task benchmark, high- and low-consequence questions are drawn from the same document images, so content alone cannot reveal which questions are costly to get wrong. In this setting, our method reduces high-stakes errors from 0.300 to 0.133 under the same total token budget, whereas a content-driven allocator performs no better than uniform allocation. Measuring how error rates change with token budget across different cost ratios, we derive an allocation frontier: uniform allocation is optimal when errors are equally costly, and token transfer toward high-consequence questions becomes increasingly beneficial as the cost gap grows. This allocation principle generalizes well across three dense vision-language benchmarks, two budget realization mechanisms (token deletion and resolution reallocation), two VLM architectures, and multiple token selection strategies. On a realistic mixed workload, consequence-sensitive allocation reduces cost-weighted error by 38% while achieving approximately 21% lower latency than full-resolution inference.

Distributed deployment of large vision foundation models often partitions a ViT backbone and exchanges intermediate token features between computing nodes, making efficient feature compression critical under bandwidth and computation constraints. Existing ViT feature codecs typically flatten heterogeneous global and patch tokens into an L x C pseudo image, causing entropy models to mainly capture sequence-axis dependencies while overlooking the native two-dimensional patch-grid structure. In this paper, we show that ViT patch tokens retain strong local spatial correlations on the original grid. To exploit this structural prior, we propose the Visual Token Codec (VTC), a dual-path learned codec that separates global and patch tokens into dedicated coding paths. Global tokens are compressed with a lightweight factorized prior, whereas patch tokens are encoded on the patch-token grid using a spatial-channel context entropy model. To support intermediate-layer compression and practical rate adaptation, VTC further incorporates feature-matching supervision after subsequent ViT blocks and variable-rate modules within a single codec. Experiments on DINOv2 and SAM3 show that VTC consistently outperforms representative ViT feature coding baselines on classification, segmentation, and detection tasks. At 90% of uncompressed-feature performance, VTC reduces bitrate by 15.7x-37.4x across these tasks. We further provide intermediate-layer rate-utility analyses for practical transmission- and storage-oriented deployment scenarios.

A single model scale challenges the flexibility of a production retrieval system: some settings need it faster, others need a smaller index, and the right trade-off changes with the workload. In the context of information retrieval (IR), a transformer-based model can be made smaller in three ways---using fewer layers, passing fewer tokens through the upper layers, or producing a shorter embedding---and each way saves a different compute resource. These options have been studied one at a time, each as its own method with its own code and training setup, which makes them hard to combine or adapt to a new model. We present~\ours to bring all three under one simple abstraction: a single object names any size the model can run at, and a short schedule lists the sizes to train. Training then produces one checkpoint that serves all of those sizes, and at deployment the user picks any of them. The same abstraction covers both retrievers and rerankers and both encoder and decoder models, as it works through interfaces that Hugging Face transformers already expose; a new backbone is a configuration change, not new modeling code. Prior methods---Matryoshka embeddings, early exit, 2DMatryoshka (e.g., Starbucks), and layerwise token compression---become special cases of our unified abstraction. The same interface also enables MatryoshkaLTC (MLTC), which jointly trains several token-compression ratios in one retriever checkpoint. To validate our framework, we train 20 checkpoints across three backbones and two tasks: the quality curves are smooth, one checkpoint costs little over a model trained for a single size, and a controlled study confirms the wallclock speedups. We release the framework and all checkpoints as a resource for building elastic retrieval systems.

3D vision-language models (3D VLMs) enable spatial reasoning over multi-view scenes but suffer from substantial token redundancy due to duplicated observations and large uninformative regions, leading to high computational cost. Although visual token compression has shown promise in accelerating 2D VLMs, it fails to capture the structured nature of 3D scenes and leads to incomplete spatial coverage and loss of fine-grained details. In this paper, we propose \textbf{HiSC}, a training-free framework for hierarchical spatial clustering token compression in 3D VLMs. HiSC lifts token compression from token-level selection to cluster-level processing by organizing tokens into spatially grounded clusters using joint geometric and semantic cues. Specifically, we first introduce a \textbf{spatial graph-based merging (SGraM) strategy} that models cross-view redundancy as spatial connectivity and consolidates physically consistent regions, effectively merging extremely similar redundant tokens prior to LLM inference. We then propose a \textbf{spatial clustering-based pruning (SCluP) paradigm} within LLM inference, which performs hierarchical compression across clusters and within clusters, preserving object instance completeness while retaining fine-grained details for important regions. Extensive experiments on diverse 3D reasoning benchmarks show validate the effectiveness of HiSC, particularly under high visual token pruning ratios. Besides, HiSC achieves over 90% token reduction with minimal performance degradation. Code is accessible at https://github.com/elecreak/HiSC.

Omni-modal large language models (Omni-LLMs) have achieved remarkable performance on audio-visual understanding tasks, but processing long and highly redundant visual and audio token sequences incurs substantial computational overhead, demanding aggressive token compression for efficient deployment. Existing methods often degrade at low token budgets: pre-LLM compression may discard structurally important and globally distributed evidence, whereas inner-LLM compression often underexploits query-conditioned audio-visual collaboration. To address these limitations, we propose OmniPack, a training-free framework that coordinates structural compression before the LLM with task-relevant semantic refinement within the LLM. Before the LLM, OmniPack removes structural redundancy through modality-specific importance, global coverage, and similarity-aware merging. After sufficient multimodal interaction, it further consolidates diverse, task-relevant representations through textual guidance and audio-visual collaboration. Extensive experiments on five benchmarks with three Omni-LLM backbones demonstrate that OmniPack consistently achieves the best performance-efficiency trade-off across diverse retention ratios, outperforming all existing methods. Notably, on Qwen2.5-Omni-7B, OmniPack preserves 98.0% of the original performance while reducing FLOPs to 16.7%, and still retains 92.9% of the original performance with only 6.8% of the original FLOPs.

Recent visual-text compression (VTC) methods, typified by DeepSeek-OCR, report impressive high token compression ratios for long-context modeling tasks by leveraging text-to-image rendering. However, existing evaluation protocols heavily rely on downstream task performance. Such evaluation metrics fail to accurately measure text preservation due to the strong inherent linguistic priors of Multimodal Large Language Models (MLLMs). In this work, we introduce a new evaluation framework that decouples MLLMs' capabilities to faithfully assess VTC quality. Within this framework, we further introduce the ZeroSense Benchmark to ensure low semantic correlation of testing samples. By eliminating textual dependencies, our benchmark guarantees that the evaluation results are purely reflective of VTC quality, unaffected by the semantic inference capabilities of downstream models. Extensive experiments across multiple datasets demonstrate that VTC quality and downstream task accuracy diverge significantly, highlighting the necessity of our decoupled evaluation framework.

Token compression in OmniLLMs is typically posed as a single saliency-ranking problem: score each multimodal token, keep the top-K. We argue this abstraction is mis-specified. The same attention score simultaneously decides two things: how much retained capacity each modality receives, and which tokens within a modality are kept. A shared top-K rule therefore inherits this audio-favoring allocation prior, spending retained capacity on audio before video tokens have a chance to compete. We propose Macer, a training-free compressor that first assigns explicit audio and video budgets, then performs allocation-normalized ranking within each modality at modality-specific shallow layers. Macer significantly reduces token cost while preserving accuracy across audio-grounded, audio--video joint, visual-dominant, and video-centric benchmarks. At 25 % retention, Macer preserves 98.7 % of full-token performance on Qwen2.5-Omni-7B and 97.3 % on Qwen2.5-Omni-3B. On Qwen2.5-Omni-7B, this 25 % setting reaches OmniZip-level performance at 45 % retention while using lower FLOPs. On OmniVinci-9B, the same allocation-before-ranking principle improves over shared top-K ranking by up to 12.9 points.

Video large language models (Video-LLMs) represent videos as dense sequences of visual tokens, whose length grows with the temporal and spatial extent of the input. These tokens often contain substantial redundancy arising from repeated visual patterns, leading to unnecessary computation in the subsequent language-model processing. Existing token compression methods, including pruning and merging, perform compression online during inference, repeatedly incurring additional computation for each input video and often relying on model-specific designs that limit their generality, we instead rethink this paradigm by shifting the costly compression process offline. We propose \textbf{ONCE}, a plug-in video token compression framework that introduces an offline-to-online paradigm: a frequency-aware global codebook is learned once in the visual feature space and reused for lightweight online compression through codebook lookup and aggregation, reducing repeated per-video computation and the need for model-specific compression designs. Extensive experiments across multiple video understanding benchmarks and against diverse compression baselines demonstrate that our approach achieves a strong accuracy-efficiency trade-off, maintaining competitive performance while achieving the lowest inference latency among compared methods.

Recent 3D vision-language models (3D VLMs) construct geometry aware tokens by projecting 2D visual features into world coordinates, enabling spatial reasoning for tasks such as 3D question answering. However, this design generates thousands of tokens per scene, resulting in substantial computational and memory overhead. While token compression has been extensively studied in 2D VLMs, existing approaches rely on semantic relevance or attention-based selection that overlook the structured spatial nature of 3D tokens. Moreover, redundancy in 3D representations cannot be resolved by spatial proximity alone, as object-level token imbalance persists even after spatial aggregation. To address this, we propose 3DZip, a three-stage token compression framework that first applies coarse voxelization to remove point-level redundancy, then selects anchor tokens based on feature-space diversity via a Determinantal Point Process, and finally merges remaining tokens under spatial constraints to preserve geometric coherence. Experiments on three 3D question answering benchmarks demonstrate that 3DZip consistently outperforms existing compression methods, retaining 94.7% of the original performance with only 128 tokens, achieving a $1.92\times$ faster inference speed.

Streaming VideoLLMs process frames causally while visual tokens grow continuously, making compression essential for controlling prefilling latency and memory. Existing training-free methods independently rank tokens, ignoring marginal-gain interactions among retained tokens. We argue that streaming video token compression should instead be formulated as set selection, where each candidate is valued by what it adds beyond the tokens already retained. Unlike existing set-wise methods designed for offline tasks, streaming makes causal, frame-by-frame pruning decisions, so modeling cross-frame interactions requires an explicit historical reference. This creates a reference-set dilemma: the reference must adequately represent previously conveyed content while remaining bounded for real-time inference. We introduce NovaCov, to our knowledge the first training-free, plug-and-play set-wise token compressor designed for streaming video. NovaCov maintains a capacity-bounded, recency-weighted Historical Reference Bank and optimizes a dual-branch submodular coverage objective that preserves representative current-frame content while prioritizing information insufficiently covered by history. Both branches are facility-location functions, so greedy selection retains the classical (1-1/e) approximation guarantee. Across streaming and offline benchmarks, NovaCov outperforms existing training-free compression methods, retaining 99.6% of ReKV accuracy while reducing LLM prefilling latency by 46%.

A 3D CT scan entering a vision-language model produces a long sequence of visual tokens, often thousands to tens of thousands per volume, and this sequence must be compressed before a language model can consume it. Token compression is well studied in general vision, but little of it targets 3D CT specifically. A common baseline is grid average, which pools regular grid cells and can blend distinct anatomy, lesion, and air into one token. We present \textbf{ORCA} (ORgan-Centroid Aggregation), a token compressor for 3D CT. It merges adjacent tokens with organ guidance and adds a sinusoidal encoding of each region's centroid to preserve spatial layout. This preserves the anatomical information a downstream model needs. ORCA is training-free and plug-and-play, producing an adjustable token set without any model change or text query. We evaluate it across two datasets (CT-RATE and Merlin) and five encoders. The evaluation spans two task types: attribute prediction over five families (size, density, location, texture, and disease) and text generation (visual question answering and report generation). At matched token budgets, ORCA improves consistently over existing compression methods. It shrinks the visual context $64\times$ and its KV-cache $50\times$, and is $31\times$ faster to process each volume. Code released at https://github.com/renjie-liang/ORCA-3DCT.

Emerging Omni-modal Large Language Models (OmniLLMs) enable unified understanding of text, audio, and video, but their long audio-video token sequences introduce substantial memory and inference costs. Existing compression methods mainly focus on selecting important tokens under fixed budgets, leaving the preceding budget-allocation problem underexplored. We show that direct query-to-audio/video similarity is unreliable for inter-modal budget allocation, and that uniform intra-modal budgets can miss key evidence while retaining redundant content. To address these limitations, we propose OmniDelta, a training-free, skill-driven framework that couples intent-aware inter-modal allocation with content-aware intra-modal allocation. OmniDelta first constructs audio and video skill pools to shift the fixed retained-token budget according to query demand, then reallocates modality budgets over audio segments and video frames using local complexity and temporal redundancy. The resulting local budgets can be combined with existing pruning strategies, preserving the total retained-token ratio while changing where the budget is spent. Experiments on four audio-video benchmarks with two Qwen2.5-Omni models show that OmniDelta establishes a new accuracy-efficiency Pareto frontier across pruning ratios. At 25% token retention on Qwen2.5-Omni-7B, OmniDelta reduces GPU memory by 22.0% and achieves a 1.64x end-to-end speedup over full-token inference.

Recent work has shown that subword vocabularies can be trained to optimize compression for a specific inference rule rather than relying on greedy heuristics such as Byte Pair Encoding (BPE). We extend this approach to greedy left-to-right longest-match decoding, the fast and widely used inference rule underlying WordPiece. We introduce Joint Optimization for Greedy Longest-Match Tokenization (JOLT), which formulates vocabulary learning as an integer program over vocabulary-selection and segmentation-choice variables. Greedy-consistency constraints ensure that each optimized segmentation exactly matches the segmentation produced by longest-match decoding under the selected vocabulary, aligning the training objective with deployment-time tokenization. To scale the optimization, we solve a linear programming relaxation and selectively introduce higher-order segmentations only for unresolved pretokens. The resulting relaxation is nearly integral: rounded solutions fall within 0.008 - 0.176 % of the LP lower bound on the training scope. The bound also shows that BPE is already within 1 - 2 % of the best achievable compression under greedy longest-match decoding, while JOLT closes 89.6 - 99.4 % of the remaining gap. On held-out validation data across four training scopes and vocabulary sizes of 32,000 and 64,000, JOLT produces up to 0.78 % fewer tokens than BPE, with improvements generally increasing as the training scope grows. These results demonstrate that inference-aligned vocabulary optimization can recover most of the limited compression headroom left by BPE while providing a certificate of near-optimality.

Existing Large Vision-Language Models (LVLMs) struggle with long-form video understanding due to the quadratic computational cost of visual tokens. While recent efficient methods attempt to compress tokens via hard pruning or uniform merging, they operate strictly in the spatial feature domain, where robust structural context and discriminative semantic details are inherently entangled. In this work, we propose WaveZip, a joint signal-frequency-domain framework for efficient video inference. Driven by the insight that temporal redundancy resides in low-pass approximation scales while spatial saliency strongly correlates with high-frequency components, WaveZip leverages Discrete Wavelet Transforms (DWT) to disentangle these signals. Temporally, it employs 1D DWT to analyze query-frame relevance, and the resulting high-frequency coefficients are further gated by inter-frame differences, with both signals jointly driving the dynamic allocation of a precise frame-level token budget. Spatially, a 2D DWT decomposes features into low-frequency approximations and high-frequency detail components, where the high-frequency coefficients are modulated within query-salient regions to regulate spatial reconstruction. Importantly, WaveZip requires no task-specific training and can be seamlessly integrated into off-the-shelf LVLMs to boost inference efficiency. Extensive experiments on long video understanding benchmarks demonstrate that WaveZip retains 99.6% of the full performance under an extreme 10x compression ratio, consistently outperforming state-of-the-art methods.

Existing token compression methods for omnimodal large language models typically rely on one modality to determine what to retain in the other. We show that this assumption often breaks down: for the same query, audio and video relevance often peaks at different moments. This cross-modal salience mismatch makes unidirectional guidance prone to discarding answer-critical cues under aggressive compression. We propose OmniScope, a training-free token compression framework that uses the query as a shared semantic anchor while estimating relevance separately for audio and video. OmniScope allocates modality-specific token budgets, prunes visual tokens with an anchor-delta strategy that preserves both global context and temporal changes, and merges audio tokens within each second to reduce redundancy while maintaining temporal continuity. Across four audio-video benchmarks and two Qwen2.5-Omni model scales, OmniScope achieves the best average accuracy across all compression settings. At 25% overall token retention, it delivers up to 3.53x prefill speedup and more than 15% GPU memory reduction, with only a 0.35-point drop in average accuracy. These results suggest a simple design principle for OmniLLM inference: share the query across modalities, but not the salience estimates. The code is available at https://github.com/MAC-AutoML/OmniScope.

Token cramming compresses sequences into learned embeddings with near-perfect reconstruction, but fixed token budgets and 99% accuracy thresholds leave it unclear whether residual errors reflect optimization failures or fundamental limits. We introduce progressive cramming, which grows the target prefix token-by-token, stopping only when reconstruction is no longer achievable within a fixed optimization budget. Progressive trajectories occupy low-dimensional structure in embedding space. Prepending a crammed embedding causes a moderate but consistent accuracy drop on multiple-choice benchmarks even with the original prefix in context, and collapses capability almost entirely under generative evaluation. Causal attention-knockout interventions trace this degradation to the embedding's interactions in the model's early layers. These results position progressive cramming as a tool for studying compression limits and show that perfect reconstruction - achievable through brittle steering rather than transferable semantics - is insufficient for meaningful compression.

The goal of this paper is to reduce the input token cost of Omni-modal large language models (Omni-LLMs) at inference time. Omni-LLMs reason jointly over audio, video and text, but the cost of the three streams is highly unbalanced: visual tokens account for the vast majority of the input, and are highly redundant. In this paper, we propose ReMo, a training-free framework that compresses visual tokens by redistributing their information across modalities: a visual token is kept only if its information appears nowhere else. ReMo achieves this in two ways: (i) it aligns audio and video in a common embedding space, and removes visual tokens already explained by the audio or by other visual tokens; and (ii) it replaces object-level visual tokens with compact text proxies, short descriptions of each object and its location, conveying the same content in far fewer tokens. On Qwen2.5-Omni at two model scales, ReMo removes 54% of the input tokens with no loss in accuracy. Indeed, it slightly exceeds the full-token model, reaching 101.2% and 101.3% of its average accuracy over five audio-visual benchmarks.

In this paper, we show for the first time that visual token pruning enhances the robustness of Multimodal Large Language Models (MLLMs), mitigating vulnerabilities such as jailbreak attacks and hallucinations. Given that vision and language modalities cannot be perfectly aligned, the misaligned visual tokens might act as out-of-distribution (OOD) inputs, leading to unpredictable outputs and introducing potential vulnerabilities. Building on this insight, we aim to enhance model robustness against jailbreaks and hallucinations by reducing OOD visual tokens at robust-pruning layers, while also reducing inference cost as a side benefit. Specifically, we measure the distance between each visual token and the language feature space. Then, visual tokens with large distances are identified as OOD tokens, which can be iteratively pruned. To demonstrate the effectiveness of our method, we evaluate it on seven diverse popular benchmarks. Notably, our method yields an average improvement of 13.29% in defending jailbreak attacks, consistently achieves competitive performance in mitigating hallucinations, and maintains strong results on general datasets like MME.

Vision-language models (VLMs) process large numbers of visual tokens, resulting in substantial inference latency and memory overhead. This has motivated extensive research on visual token compression. While training-free strategies rely on heuristic metrics and suffer significant performance degradation under high compression ratios, many training-based methods introduce external compression modules that force the VLM backbone to adapt, incurring substantial retraining cost and compromising VLMs' priors. Effective visual token compression hinges on strong information encoding, a capability already present in pretrained VLMs but underutilized by existing approaches. Motivated by this, we propose VisCo, a training-efficient self-compression framework that reuses the pretrained VLM itself as an intrinsic compressor. VisCo is a parameter-sharing autoencoder that compresses visual information using a small set of memory tokens and transfers hierarchical information from encoding to decoding. Experiments show that VisCo surpasses prior methods across all evaluated compression ratios, with larger gains under more aggressive compression, and remains stable even in the extreme single-token setting. Moreover, when combined with the original visual tokens, the learned memory tokens can even improve the base model, suggesting that VisCo captures complementary representations beyond compression.

Although Video Large Language Models (Video LLMs) have shown strong performance in video understanding, their efficiency is still limited by the large number of visual tokens. Existing video token compression methods typically rely on frame-wise saliency or heuristic token merging, which can over-focus on locally salient regions and produce ambiguous fused features. To address these issues, we propose GeoTrace, a training-free spatiotemporal token compression framework that decomposes video evidence into exact skeleton tokens and traceable residual event tokens. Specifically, Contextual Farthest-Point Anchoring (CFPA) preserves salient, context-consistent, and high-coverage skeleton tokens, while Trajectory-Constrained Residual Condensation (TCRC) compresses residual tokens through one-to-one temporal trajectories and constrained near-manifold condensation, producing traceable event tokens with reduced ambiguity. We evaluate GeoTrace on four Video LLMs across four video understanding benchmarks, and the results demonstrate its effectiveness and generalization across different model architectures and scenarios. On LLaVA-OneVision, with only 10% visual tokens retained, GeoTrace achieves a $12.99\times$ TFLOPs reduction while preserving 99.1% of the vanilla performance. Overall, GeoTrace offers a compact and traceable token representation for efficient and robust Video LLM inference. Code is available at \href{https://github.com/guohuan-xie/GeoTrace.git}{\texttt{Code}}.

Omni modal large language models (OmniLLMs) have attracted wide attention for their ability to jointly process audio and video, but they generate large token sequences under audio-visual inputs, leading to substantial inference cost. Existing audio-visual token compression methods often rely on unimodal guidance, overlooking the temporal locality of query-relevant evidence in audio-visual inputs and implicitly assuming that the two modalities share a temporally aligned information density distribution. We propose \textbf{OmniFocus}, a training-free query-guided token compression method for OmniLLMs that performs independent importance estimation for video and audio, enabling a modality-symmetric compression design that preserves modality-specific salient evidence while maintaining audio-visual alignment, thereby mitigating the modality bias issue that can arise from unimodal-guided compression. Experiments on the Qwen2.5-Omni model family across four audio-visual benchmarks show that OmniFocus maintains strong compressed performance at low token retention ratios and outperforms existing baselines on several major benchmark scores at 25% token retention. On DailyOmni with Qwen2.5-Omni-7B at 25% token retention, OmniFocus maintains 59.40 accuracy while delivering up to 1.38$\times$ prefill speedup relative to the full-token baseline, highlighting a favorable practical accuracy-efficiency trade-off.

Video reasoning segmentation demands pixel-accurate object tracking across hundreds of frames under complex natural language queries, producing dense spatiotemporal tokens whose quadratic self-attention cost makes long-video processing prohibitive. Existing methods address this through token compression, yet typically operate on encoder features lacking temporal context, constraining selection before content redundancy can be reliably assessed. Informed compression requires contextual awareness, but acquiring that awareness at full resolution incurs the same quadratic cost compression aims to reduce. State-space models resolve this constraint, as their linear recurrence selectively conditions each token on temporal context at $\mathcal{O}(T)$ cost, producing representations where content redundancy becomes assessable. Building on this, Selective SpatioTemporal Aggregation and Compression (STAC) enriches features via decoupled bidirectional spatial and causal temporal scanning, leveraging recurrence-derived redundancy for hierarchical compression with adaptive thresholds optimised with segmentation objective. STAC achieves 85% token reduction and 1.8$\times$ speedup while surpassing compression-free baselines on reasoning segmentation benchmarks in a zero-shot streaming-compatible setting. Code is available \href{https://github.com/MCG-NKU/nku-video}{here}.

Vision Transformers (ViTs) are strong backbones for semantic segmentation, but their computational cost limits deployment. Recent token compression methods for efficient transformer-based segmentation reduce this cost by decreasing the number of tokens. However, existing evaluations primarily focus on low-to-moderate compression, leaving their behavior under aggressive compression and corrupted inputs unclear. Meanwhile, structural pruning provides an orthogonal route to efficiency by removing redundant components in the ViT architecture, but is rarely compared to token compression under a unified protocol. To bridge this gap, we benchmark representative token compression and structural pruning methods for ViT-based semantic segmentation under matched FLOPs on ADE20K and Cityscapes, together with their common-corruption variants ADE20K-C and Cityscapes-C. Our results reveal a consistent trend on both clean and corrupted inputs: token compression is highly effective at mild reductions but degrades sharply when compression becomes severe, consistent with substantial information loss from overly aggressive token reduction. In contrast, structural pruning exhibits a smoother degradation curve and is more stable at high compression. Motivated by these findings, we study a prune-then-merge pipeline that applies moderate token compression on top of a moderately pruned backbone. At comparable FLOPs, this combined strategy consistently achieves a better accuracy-robustness trade-off at high compression, offering a practical recipe for deployment-oriented ViT segmentation. Code is available at https://github.com/phatnguyencs/vit-seg-compression.

The Unigram tokenizer uses an elegant representation which makes it straightforward to edit vocabularies, but its training is comparatively heavy and complex. We introduce MinGram (Minimalist Unigram), which keeps the token-list representation but simplifies training using a BPE-derived seed vocabulary, Hard EM on a minimum-token path, and a single flat score-pruning step. This removes the suffix array, the forward-backward pass, and the iterative prune loop, leaving a procedure that requires little beyond tokenizer inference itself. By making token count the primary objective and using a Unigram score only as a tiebreak, MinGram keeps the compression of pure token-count methods while retaining much of the morphological alignment and downstream quality of probabilistic ones. Across six languages, MinGram compresses better than both BPE and standard Unigram, and a compression-oriented variant matches the strongest token-count compressors while retaining substantially higher morphological alignment. In controlled downstream language-model training, Unigram-family tokenizers, with MinGram among the best, consistently beat BPE in bits-per-byte.

High-precision remote perception is often hindered by the severe bandwidth constraints of Vehicle-to-Everything (V2X) networks. We propose \textit{DinoLink}, a token-centric compression framework that replaces raw pixel streaming with discrete semantic communication for vehicle-cloud collaborative inference. DinoLink employs a dual-sparsity architecture: a saliency-aware selector prunes redundant background tokens, while a Residual Vector Quantization (RVQ) module collapses features into compact codebook indices. By transmitting only lightweight indices and positional priors, DinoLink achieves a $139\times$ bitrate reduction compared to uncompressed transmission while maintaining a competitive 32.8% mAP on the nuScenes dataset. Deployment simulations further demonstrate a $34.5\times$ acceleration in narrow-band environments, such as LoRa. Our results substantiate DinoLink as a robust, bandwidth-efficient frontend for high-fidelity remote perception in constrained V2X scenarios. The code is publicly available at https://github.com/UGA-MOBILITY-LAB/dino_link.

Multimodal Large Language Models have achieved remarkable progress in short-form audio-video understanding, yet long-form audio-video comprehension remains challenged by limited context windows and severe information redundancy. To address these bottlenecks, we propose AVOC, a framework for long-form audio-video understanding in Omni-modal Large Language Models. AVOC introduces a learnable token compression module between the modality encoders and the LLM backbone. We reframe multimodal token compression as a top-$K$ retrieval problem: given a fixed context budget, the module must retrieve a compact subset of tokens that best supports answering the user query. We draw inspiration from three classical Information Retrieval criteria for selecting informative units from a large candidate pool: relevance, importance, and diversity. AVOC instantiates each criterion as a tailored mechanism for audio-video understanding, and integrates them into a unified retrieval-style compression pipeline. Experiments show that AVOC achieves state-of-the-art performance on long-form audio-video benchmarks, surpassing the second-best model by 4.9 and 5.5 points in average accuracy on OmniVideoBench and LVOmniBench, respectively. Moreover, AVOC maintains robust performance on Audio-Video Needle-in-a-Haystack task at durations up to one hour.

Deploying vision transformers (ViTs) on sensor-edge systems is limited not only by on-device compute, but also by the energy and bandwidth required to transmit high-dimensional image data from the sensor to the processor. While in-sensor and near-sensor computing reduce this cost through early feature extraction, existing methods often provide only modest compression. We observe that the frequency domain provides a naturally compact representation of visual information and can be exploited at the sensor level to reduce sensor-to-processor data movement. Building on this insight, we present FrequencyFormer, a co-designed sensor-to-processor pipeline for efficient ViT inference. FrequencyFormer includes: (1) a multi-scale DCT tokenizer that compresses a 224x224 image into compact frequency-domain tokens, achieving up to 128x reduction in off-chip data volume with modest accuracy loss; (2) a LUT-based near-sensor hardware implementation that leverages fixed DCT coefficients for multiplier-free, energy- and area-efficient tokenization; and (3) a modified MIPI-based low-power communication architecture that further reduces transfer energy. FrequencyFormer serves as a drop-in replacement for standard ViT patch embedding and remains compatible with pretrained backbones across classification, detection, and segmentation tasks. The pipeline achieves 28.8 TOPS/W, reduces communication energy by 230x, and lowers total sensor-side energy by 2.22x, demonstrating frequency-domain tokenization as a scalable foundation for in-sensor ViT deployment.

Personalized dialogue agents require continuous long-term memory to maintain coherent interactions across multiple sessions. However, deploying these capabilities on consumer-grade hardware (e.g., 8 GB VRAM edge devices) introduces severe memory and compute bottlenecks. Existing systems typically rely on isotropic cosine similarity for retrieval and heuristic rules for context compression. These approaches lack a unified theoretical foundation, frequently suffering from the hubness problem in high-dimensional retrieval and syntactic fragmentation during compression. To overcome these limitations, we propose CoreMem, a resource-efficient edge-cloud memory architecture fundamentally unified by information geometry. First, Riemannian retrieval replaces cosine matching with a locally adaptive Fisher-Rao metric, effectively penalizing hub memories via Mahalanobis distance with O(Ndr) Woodbury acceleration for real-time search. Second, Fisher-guided discrete token distillation (FDTD) introduces a hierarchical sentence-to-token compression mechanism. It derives sensitivity scores from Fisher information traces, providing a principled compression-KL tradeoff augmented with explicit structural syntax protection. Evaluated on the LOCOMO and LongMemEval-S benchmarks, CoreMem achieves strong accuracy improvements, yielding substantial gains in Open-domain (+4.51 pp) and Temporal (+4.17 pp) reasoning. Extensive profiling confirms that CoreMem operates seamlessly within a strict 8 GB VRAM budget, successfully bridging the gap between resource-constrained edge devices and the demand for theoretically grounded, lifelong memory agents.

Video tokenization is fundamental to scalable video generation, as the number of tokens directly determines the computational cost and the length of videos that can be modeled. Existing tokenizers mainly improve scalability by compressing videos into fewer tokens, but they often continue to represent persistent content, such as static backgrounds and consistent object appearances, repeatedly across frames and chunks. In this paper, we propose \textbf{TivTok} (\textit{Time-Invariant Tokenizer}), a reuse-aware video tokenizer that makes persistent information reusable across time. TivTok represents a clip with Time-Invariant (TIV) tokens that encode information shared across frames and Time-Variant (TV) tokens that encode frame-specific residuals. To obtain this factorization, we introduce Scope-Induced Factorization (SIF), which assigns different attention scopes to the two token groups: TIV tokens attend to the full clip, whereas each TV token only accesses its corresponding frame together with the TIV tokens. In the decoder, Invariant Broadcasting (IB) reuses the same TIV tokens across frames and chunks for parallel reconstruction and long-video tokenization. Experiments show that TivTok achieves an rFVD of 12.65 on the standard $16{\times}256{\times}256$ benchmark and improves compression efficiency by 2.91$\times$ for 128-frame videos compared with the evaluated baselines, while using only 1.1% of the tokens required by downsample-based tokenizers in our evaluation.

Creative image editing tools, such as Photoshop's Remove or Generative Fill buttons, are central to everyday customer use and account for a major share of traffic in Photoshop and Lightroom. However, current generative AI models face significant latency challenges, which become even more pronounced when transitioning from convolution-based U-Nets to Diffusion Transformers (DiTs). In our evaluation on hundreds of representative image editing samples spanning a wide range of mask ratios, the DiT module alone accounts for an average of 73% of the total model latency, even after being distilled from 50 timesteps down to 8 timesteps. To tackle this challenge, we propose $\textbf{HiLo-Token}$, an input-adaptive token compression framework that allocates more token budget to high-frequency, rich-context regions while assigning fewer tokens to low-frequency areas. Specifically, for the editing region specified by the user mask, we retain all tokens within a dilated mask to preserve strong locality and contextual relevance. Outside the editing region, we introduce a simple yet effective high-frequency token selection strategy based on spatial frequency to capture important local details, while using tokens from a 16x downsampled image to represent low-frequency components and preserve the blurry but global structure. Extensive experiments on production-level evaluation data validate the effectiveness of the proposed method, achieving 3.13x, 2.59x, and 1.67x DiT speedups on A100-80GB for image editing tasks across small, medium, and large mask ratio categories with average ratios of 6.38%, 15.92%, and 35.36%, respectively, without any regression in generation quality.

Large language models (LLMs) have enabled time series (TS) analysis by jointly modeling numerical observations and textual context through a shared token interface. However, TS tokens and prompt tokens exhibit fundamentally different information structures, making uniform token processing inefficient. In this paper, we study token efficiency in TS language modeling from an asymmetric-token perspective. We show that TS tokens have highly uneven spectral contributions, where many tokens share redundant frequency patterns while a small subset preserves critical temporal evidence. We also observe that prompt-token influence attenuates with model depth, suggesting that full prompt retention across all layers is unnecessary. Based on these findings, we develop an adaptive token budgeting framework that compresses TS tokens via frequency-domain structure and progressively reduces prompt tokens across layers. Experiments across forecasting, classification, imputation, and anomaly detection demonstrate up to \textit{\textbf{7.68$\times$}} inference acceleration and performance gains in \textit{\textbf{78%}} of evaluated settings, showing the effectiveness of asymmetric token compression for scalable TS foundation models.

Large language models (LLMs) are widely used to tackle complex tasks with autonomous workflows. Recently, reusable natural language skills have emerged as a popular paradigm to inject procedural knowledge into LLM applications. Since popular skills are often invoked repeatedly, placing their full text in every context significantly increases prefill cost and latency. While text compression techniques have the potential to solve this problem, most existing methods are designed to compress factual knowledge in documents instead of procedural knowledge, making them insufficient for skill compression. In this paper, we argue that an effective skill compression method should: 1) preserve logical dependencies among workflows and tool protocols, 2) enable lightweight, offline compression for frequently updated community skills, and 3) be adaptable to varying complexities across skills. To address this, we present SKIM (SKIll coMpression), an adaptive multi-resolution soft token compression framework for procedural skills. Depending on the complexity of each skill, SKIM creates different numbers of soft tokens that not only improve the efficiency of LLM inference, but also preserve the effectiveness of skill usage. Experiments indicate that SKIM compresses skills to 30 to 60 percent of their original token length while preserving task performance better than existing compression methods.We have released our code at https://github.com/bebr2/SKIM .

Monolithic vision-action models represent an emerging paradigm in autonomous driving. However, this architecture produces token sequences that quickly exceed real-time computational budgets when encoding extended temporal context for complex interactions. While approaches like linear transformers and external memory try to make the context lightweight, token compression is most compatible with the architecture as it requires no backbone modifications. Yet existing compression adopts rule-based heuristics like temporal decay, decoupled from planning, risking loss of decision-critical information. We propose COMPACT-VA, a planning-aligned working memory framework built on conditional VQ-VAE, compressing extended context into bounded representations. Compression is conditioned on both historical trajectory and a learned planning intent that the posterior encoder distills from future trajectories during training, while the prior encoder learns to predict it from compressed observations. The compressed memory, concatenated with the predicted latent, feeds the policy for end-to-end optimization, planning with retained decision-critical information. We evaluate on high-signal dynamic scenarios where historical context is most critical for behavior correctness (e.g., stop, yield, or proceed), and accordingly design behavioral metrics. Under comparable token budgets, we achieve $>$6% improvement (68.3%) on success rates with consistent gains across metrics. Ablations validate planning-aligned coupling effectiveness. Closed-loop evaluation confirms that COMPACT-VA maintained general driving performance with 3.3* speedup and 2.7* memory reduction over uncompressed processing.

Adaptive video tokenisation seeks to dynamically allocate token budgets based on the underlying visual complexity of a sequence. Current continuous-regime approaches achieve this via iterative binarised searches or trained neural regressors, while discrete methods often require a full-rate decoder pass to estimate information content. We demonstrate that such computational overheads are not strictly necessary. We show that the latent space of a frozen continuous video tokeniser inherently encodes temporal redundancy that can be exploited directly: spatial positions whose latent representations change minimally between consecutive frames carry near-zero additional information. We introduce a parameter-free adaptive token allocation mechanism that applies a fixed threshold to per-position temporal-L1 differences, identifying and dropping redundant latent positions. Consequently, the compression rate emerges naturally from the input content rather than being enforced top-down: static scenes get compressed aggressively, while highly dynamic sequences retain more tokens. To reconstruct the dropped positions, we propose the Latent Inpainting Transformer (LIT), a lightweight factorised spatial-temporal attention architecture. The resulting inference pipeline is highly efficient, requiring only a single encoder pass and one LIT forward pass, eliminating the need for auxiliary routing networks. Evaluations across TokenBench and DAVIS, which are the standard benchmarks used by recent tokenisers~\cite{infotok, agarwal2025cosmos}, indicate that our framework yields meaningful, content-driven token allocation while maintaining competitive reconstruction fidelity, and delivers a $31\times$ inference-time speedup over the continuous adaptive baseline (ElasticTok-CV) and an $\approx2\times$ speedup over the discrete information-theoretic baseline (InfoTok)

LLM pipelines waste substantial token budgets on low-information content: repeated context, verbose responses, and redundant boilerplate. We introduce Entropy Gate, a token compression framework applying entropy quenching $-$ a thermodynamic process that progressively freezes out low-energy tokens while preserving semantic fidelity. Each token receives a multi-factor information energy $E(t)$ combining statistical, structural, and positional components. An adaptive quenching schedule $T(τ) = T_0 / (1 + ατ)$ removes tokens whose Boltzmann survival probability $p_i = \exp(-E_i / kT)$ falls below threshold, with a fidelity gate halting compression when energy-weighted similarity drops below $θ$. We prove token selection by descending $E(t)$ maximizes expected semantic preservation, that quenching produces nested survival sets, and that achievable compression approaches the information-theoretic limit $\text{CR} \to 1 - I(P; T)/H(P)$. A Phase 1 heuristic achieves 40-60% compression across five prompt categories while maintaining $S_E > 0.80$, with energy-squared amplification $E \to E^2$ adding 10-25 percentage points. Context deduplication adds 50-70% savings on repeated blocks. Output-side quenching, motivated by findings that brevity improves accuracy, further reduces response overhead. Combined with external memory, reduction composes multiplicatively to 88-96% for agentic workloads. The framework is stateless, model-agnostic, and deploys as an OpenAI-compatible HTTP proxy.

Video Large Language Models (Video-LLMs) achieve strong performance in video understanding, but their excessive visual tokens bring substantial computational overhead. Existing training-free compression methods improve inference efficiency by reducing visual tokens, yet they often rely on local adjacent-frame similarity for temporal redundancy estimation or allocate token budgets mainly according to segment length. Such designs are sensitive to frame-level noise and fail to capture the non-uniform information distribution of real-world videos. To address these challenges, we propose InfoMerge, a training-free visual token compression method that improves token utilization through robust redundancy estimation and content-aware budget allocation. Specifically, we propose the Temporal Fingerprint Difference: a segment-level second-order temporal redundancy estimation strategy, which models the temporal similarity structure of tokens at the same spatial positions within each segment. We further introduce Content-Aware Budget Allocation (CABA), which dynamically allocates segment-level token budgets based on segment uniqueness and spectral-entropy-based representational richness. By reducing repeated preservation of redundant static regions and allocating more tokens to informative segments, InfoMerge makes better use of the limited token budget while maintaining strong performance. Extensive experiments show that InfoMerge achieves strong efficiency--accuracy trade-offs across multiple benchmarks and backbones, with more pronounced advantages under aggressive compression. On LLaVA-OneVision-7B, InfoMerge retains 98.8% of the original average performance while reducing 85% of visual tokens and achieving a 4.24-fold speedup in the prefill stage.

Large vision-language models (LVLMs) achieve strong performance on image and video understanding tasks, but their inference efficiency is constrained by the large number of visual tokens produced by vision encoders. Most existing visual token compression methods estimate token importance from attention scores or representation properties at specific layers, overlooking how visual tokens evolve across the vision encoder. Such layer-specific criteria may provide incomplete importance estimates and limit performance preservation after compression. To address this issue, we analyze layer-wise visual token evolution directions and observe that tokens form multiple group evolution directions across vision-encoder layers. Our analysis further shows that informative tokens tend to exhibit persistent deviations from common group evolution directions. Based on this observation, we propose EvoCut, a training-free and attention-free visual token compression method that estimates token importance from multi-layer evolution deviation. Experimental results show that EvoCut can retain only 11.1% of the visual tokens on LLaVA-1.5-7B while preserving 94.4% of the average performance, demonstrating its effectiveness in balancing efficiency and accuracy.

In Vision-Language Models (VLMs), high-resolution images produce a large number of visual tokens, resulting in high computational costs and KV-cache overhead during inference. To address this problem, we propose an Extreme Token Compression (ETC) framework that minimizes task loss when reducing the number of input tokens based on the principle of variational information distillation. Specifically, from an information-theoretic perspective, we show that minimizing task loss requires the compact representation to preserve the instruction-aware sufficient statistic of the task-relevant visual information for prediction. In practice, ETC leverages text-to-image cross-attention to weight the original visual features to approximate the latent instruction-aware predictive statistic. Moreover, ETC introduces a variational information distillation, enabling the compact representation to preserve the essential information to recover this predictive statistic. Experiments on LLaVA-1.5-7B and Qwen3-VL-2B show that ETC remains effective even under single-token compression, substantially reducing KV-cache overhead while retaining strong task performance.

Large Vision-Language Models (LVLMs) map visual inputs into dense token sequences, imposing a quadratic computational bottleneck for inference. Elastic visual-token compression addresses this by training a single model that can run at multiple visual-token budgets. However, existing approaches struggle under aggressive compression. Spatial-only compression, as in nested pooling, behaves as an imperfect low-pass filter and induces spectral aliasing that obscures fine-grained detail. Query-only compression, as in nested query resampling, replaces explicit grid-aligned tokens with non-local summaries and substantially degrades spatial grounding. To resolve this representational conflict, we introduce PARCEL (Pool-Anchored Resampling with Conditioned Elastic Queries for Efficient Vision-Language Understanding), a visual tokenization architecture that dynamically partitions the labor of feature extraction. PARCEL establishes spatial pool tokens as low-frequency layout anchors and conditions elastic query tokens on these anchors through Pool-Conditioned Query Resampling. This encourages query tokens to focus on complementary visual features rather than redundant spatial mapping. Extensive evaluations across 27 benchmarks show that PARCEL improves the performance-efficiency Pareto frontier, consistently outperforming existing matryoshka baselines across visual-token budgets while preserving the "train once, deploy anywhere" paradigm.

Tokenization-free hierarchical models are emerging as a promising alternative to traditional Large Language Models (LLMs), addressing inherent preprocessing issues such as vocabulary design complexity, out-of-vocabulary (OOV) errors, and language-specific constraints. However, a significant challenge in these byte-level methods is the optimization of the compression ratio, a critical factor that dictates model performance for processing bytes data via chunks. In this paper, we propose Adaptive Targeted Dynamic Chunking (ATDC), a novel byte-compression control mechanism designed to enhance the effectiveness of dynamic chunking within hierarchical architectures. Our approach utilizes curriculum learning to progressively adjust the compression ratio during training, transitioning from low to high compression to stabilize the learning process. We provide an analysis establishing the relationship between the target compression ratio and Bytes-Per-Innermost-Chunk (BPIC), allowing for tracking of chunk-size evolution throughout the training phase. Evaluations conducted on the FineWeb-Edu 100B dataset demonstrate that hierarchical models equipped with ATDC achieve competitive Bits-Per-Byte (BPB) performance compared to conventional baselines operating at both byte and token levels. Furthermore, the proposed method exhibits more stable training dynamics and superior final performance across diverse downstream tasks compared to models using fixed compression ratios, while maintaining the inherent robustness and flexibility of byte-level processing.

Video large language models (Video-LLMs) have demonstrated strong capabilities in video understanding tasks. However, their practical deployment is still hindered by the inefficiency introduced by processing massive amounts of visual tokens. Although recent approaches achieve extremely low token retention ratios while maintaining accuracy comparable to full-token baselines, most of them perform compression only at the late stage of prefilling, leaving the efficiency of the vision encoder unoptimized. In this paper, we first show that vision encoding contributes a large portion to the time-to-first-token (TTFT). Therefore, instead of compressing visual tokens only after the vision encoder, performing compression inside the encoder still leaves substantial room for exploration. Based on this insight, we propose EarlyTom, a training-free token compression framework that performs early-stage visual token compression inside the vision encoder, enabling significantly better TTFT reduction and higher throughput. In addition, we introduce a decoupled spatial token selection strategy that improves the overall compression effectiveness. EarlyTom reduces TTFT by up to 2.65x and FLOPs by up to 61% on a single NVIDIA A100 GPU for the LLaVA-OneVision-7B model, while maintaining accuracy comparable to the full-token baseline. These improvements substantially enhance the practicality of deploying Video-LLMs in real-world production scenarios.

To guarantee that an LLM's outputs conform to a specified structure, context-free grammar (CFG) decoding engines force the selection of next tokens that produce strings that conform to a given CFG. While current CFG-constrained decoding engines are highly optimized, the inherent costs arising from the massive per-step search space -- i.e. the entire token vocabulary -- result in intractably high overhead for more complex CFGs: precisely the situation where CFG engines are most useful. In this paper, we introduce CFGzip, an offline technique for compressing the token search space, which massively reduces CFG engine overhead. In experiments, we report latency reduction of up to two orders of magnitude when CFGzip is used with a SoTA grammar engine, yielding an up to 7.5x speedup in total constrained generation time: with CFGzip, constrained decoding is now feasible at scale for complex CFGs.

High-capacity watermarking is necessary for 3D Gaussian Splatting (3DGS) assets to embed rich information (e.g., ownership, provenance, and authentication codes), enabling reliable identification and integrity verification in large-scale 3D asset pipelines. Existing bit-to-token watermarking methods based on a pre-trained text encoder are limited to 77-bit messages due to CLIP's fixed 77-token context length, as tokens beyond this limit are unsupported by learned positional embeddings. To address this limitation, we introduce BitC-3DGS, a bit-compression framework that encodes multiple message bits per token. It employs a bit-compressed tokenization scheme that encodes multiple bits within the same chunk into a single semantic token. To enable recovery of the compressed information, it further introduces a dual-branch architecture for joint chunk decompression and bit decoding, along with a hard-message sampling strategy to improve combinatorial coverage during decoder training. Extensive experiments on the Blender and LLFF datasets demonstrate the effectiveness of BitC-3DGS for high-capacity watermarking, achieving high message recovery accuracy and rendering fidelity. For example, it supports 128-bit message capacity with recovery accuracy comparable to that of 64-bit messages in recent state-of-the-art methods.

The token-level extractive compressors widely used for general LM context are structurally inappropriate for LLM agents: across 17 (env, backbone, method) cells spanning two independent token-level method families, every cell collapses to mean reward <= 0.05 despite 1.3-13.3x realized compression. We name and characterize this failure mode as action-grammar destruction -- the tokens carrying action semantics (identifiers, brackets, action verbs) are exactly those self-information ranks lowest, so a general-purpose compressor reliably removes them and the environment rejects the residual. The diagnosis points to step-granularity compression. We introduce AGORA, an inference-free step-level compressor combining a structural prompt parser, an always-keep floor for format- and recency-critical content, and a 125M-parameter relevance scorer trained on counterfactual next-action-change labels (~2ms/step, zero per-step LLM toll). Across the compared inference-free and LLM-based methods, AGORA is the only one retaining >= 75% uncompressed performance in 8 of 9 cells (with the lone exception at 73%); a four-way component ablation isolates the structural floor as the dominant quality lever and the learned scorer as the source of 1.0-11.5x adaptive end-to-end compression from a single fixed keep ratio.

Audio-language models (ALMs) are increasingly used for audio captioning, question answering, and open-ended audio understanding, but their inference cost remains high when audio inputs are represented as long prefix-token sequences. These audio prefixes consume context budget, increase memory usage, and make deployment harder in resource-constrained or latency-sensitive settings. Existing training-free audio-token reduction methods mainly rely on fixed pooling or score-based pruning. Fixed pooling is content-agnostic, while score-based pruning can preserve isolated salient tokens but discard nearby acoustic context. We propose Local Temporal Bipartite Merging (LTBM), a training-free encoder-space compression method that merges similar nearby audio tokens under an explicit temporal window constraint. Beyond introducing LTBM, we use a controlled Global Merge variant to isolate whether temporal locality itself is a useful inductive bias for audio-token compression. Experiments on AudioCaps, Clotho, and MMAU with Qwen2-Audio show evidence of a task-dependent locality effect: locality-aware merging is more favorable for captioning at several compression settings, especially under stronger compression, while global matching is more competitive for multiple-choice audio understanding. A cross-backbone validation on Audio Flamingo 3 further supports the captioning-side advantage of locality-aware merging under moderate and aggressive compression.

Real-world image super-resolution aims to recover high-quality images from complex and unknown real-world degradations. However, existing generative Real-ISR methods largely inherit the dense latent representations and quadratic-cost global modeling paradigm developed for high-resolution image synthesis, causing computation, memory usage, and inference latency to scale unfavorably with resolution and thus limiting practical deployment. We argue that the key bottleneck lies not in insufficient restoration priors, but in excessive token redundancy and costly token interactions during high-resolution restoration. Motivated by this observation, we revisit Real-ISR from the perspectives of compact latent representation and linear-complexity modeling, and propose SANA-SR, an efficient one-step restoration framework. Specifically, SANA-SR employs a deep compression autoencoder with a 32x compression ratio to drastically reduce latent tokens while preserving restoration-relevant structures and textures. On top of this compact latent space, we introduce a linear-attention DiT with LoRA fine-tuning, enabling efficient high-resolution restoration with linear-complexity token mixing. Extensive experiments on all benchmark datasets demonstrate that SANA-SR achieves highly competitive and often superior quantitative performance against existing methods, while restoring clearer and more realistic textures. Moreover, after pruning, the deployed model runs in 0.019s with 407.95G MACs and 344M parameters, highlighting its strong potential for practical mobile deployment.

Recent advances in Video Large Language Models (Video-LLMs) have greatly expanded multimodal reasoning capabilities. However, the massive number of visual tokens extracted from long video sequences incurs prohibitive computational costs, limiting their deployment in real-world scenarios. Existing training-free token compression methods select tokens based on attention magnitude as a proxy for semantic importance, but often overlook positional bias and rely only on short-term temporal locality, leading to redundant spatio-temporal coverage and inefficient token usage. We present DynaTok, a training-free, temporally adaptive and bias-aware token compression framework that allocates token budgets across both temporal and spatial dimensions. Through a lightweight exponential moving average (EMA) memory, the Temporal Budget Allocation (TBA) module dynamically assigns fewer tokens to redundant frames and more to novel frames, capturing long-term temporal variation. The Spatial Budget Allocation (SBA) module complements this by selecting spatially diverse and semantically important features using activation-based attention maps, while leveraging a spatial memory to reduce redundancy from previously selected regions and mitigate positional bias. DynaTok integrates seamlessly with existing Video-LLMs such as LLaVA-OneVision and LLaVA-Video without retraining, and effectively preserves semantic coverage under aggressive compression. Experiments on four representative VideoQA benchmarks-MVBench, LongVideoBench, MLVU, and VideoMME-show that DynaTok retains over 95% of baseline accuracy even with a 90% token reduction, surpassing recent training-free approaches. These results demonstrate that DynaTok provides a principled foundation for efficient and robust video reasoning, paving the way toward real-time streaming video understanding with future Video-LLMs.

Segment Anything Model 2 (SAM 2) serves as a core foundation model in the field of video segmentation. Building upon the original SAM model, it introduces a memory bank mechanism and demonstrates outstanding performance in tasks such as semi-supervised video object segmentation and tracking anything. However, the complex computational characteristics of SAM 2's multi-stage image encoder and memory module have raised the barrier to the model's deployment in practical applications. To address this issue, we propose TinySAM 2, a lightweight video segmentation model that balances performance and efficiency. First, a memory quality management mechanism is introduced to select and retain high-informative historical frames as the memory. In addition, a joint-spatial-temporal token compression is proposed that reduces the memory storage and computational cost. Specifically, average pooling is employed to first compress redundancy tokens in the spatial domain. In the temporal domain, informative tokens are selected across frames in the memory bank based on token-level similarity measurement. Besides, we take RepViT as the lightweight image encoder, which further reduces the model parameters. Extensive experiments on challenging datasets such as DAVIS and SA-V demonstrate that TinySAM 2 achieves 90% of the performance of SAM 2.1, with only 7% memory tokens and 3% training data. This study effectively alleviates the bottlenecks in parameter count, computational load, and deployment costs associated with SAM 2, providing a resource-efficient solution for the widespread application of video segmentation models on devices.

The Segment Anything Model (SAM) achieves strong open-vocabulary segmentation, but its ViT-based image encoders dominate inference latency and memory. Existing activation compression methods, such as token merging, reduce the token length to process, yet introduce non-trivial runtime overhead and encounter catastrophic quality drop under high compression. Other methods applying Sparse Attention focus on attention alone, leaving the MLP fully dense and capping achievable speedup. We propose SparseSAM, a (i) training-free structured sparsification framework that jointly accelerates attention and MLP layers while preserving token identity. SparseSAM introduces (ii) Stripe-Sort Attention, which uses a deterministic Z-order permutation to transform dense attention into static hardware-friendly sparse patterns, eliminating dynamic masking overhead. SparseSAM further introduces a (iii) Residual-Consistency MLP that routes only informative tokens through the MLP while propagating remaining tokens through the residual pathway. Across four segmentation benchmarks, SparseSAM loses only 0.004 mIoU at a 0.4 density and 0.021 mIoU at 0.3, a 2.10x reduction in accuracy loss versus token merging advances, while achieving 2x faster inference and 2.8x memory reduction.

Adapting large AI models (LAMs) to personalized edge data is challenging because wireless devices have limited memory, computation, and uplink capacity. Federated fine-tuning preserves data privacy but still requires each device to host the full model, while split learning reduces device memory at the cost of heavy activation transmission. This paper proposes TSFLora, a token-compressed split fine-tuning framework for communication-efficient LAM adaptation at the edge. TSFLora combines attention-guided token selection, token merging, low-bit activation quantization, and LoRA-based adaptation within a split federated training pipeline. The key idea is to compress the intermediate token sequence before transmission so that the system reduces both uplink traffic and server-side processing without changing the frozen backbone. Experiments on ViT models over CIFAR-10, CIFAR-100, and TinyImageNet show that TSFLora achieves up to \textbf{6.8$\times$} communication reduction and \textbf{41%} memory saving while maintaining competitive accuracy.

Omni-modal large language models have demonstrated remarkable potential in holistic multimodal understanding; however, the token explosion caused by high-resolution audio and video inputs remains a critical bottleneck for real-time applications and long-form reasoning. Existing omni-modal token compression methods typically prune tokens at the input embedding level, relying on audio-video similarity or temporal co-occurrence as proxies for semantic relevance. In practice, such assumptions are often unreliable. To address this limitation, we propose OmniDrop, a training-free, layer-wise token pruning framework that progressively prunes audiovisual tokens within the LLM decoder layers rather than at the input-level, allowing early layers to preserve sufficient omni-modal information fusion before aggressively removing tokens in deeper layers. We further utilize text queries as guidance for modality-agnostic and task-adaptive token pruning. We also introduce a temporal diversity score that encourages balanced token survival to preserve global temporal context. Experimental results across various audiovisual benchmarks demonstrate that OmniDrop outperforms all baselines by up to 3.58 points while reducing prefill latency by up to 40% and memory usage by up to 14.7%.

Streaming video understanding with large vision-language models (VLMs) requires a compact memory that can support future reasoning over an ever-growing visual history. A common solution is to compress the key-value (KV) cache, but existing streaming methods typically rely on local token-wise heuristics, such as recency, temporal redundancy, or saliency, which do not explicitly optimize whether the retained cache is representative of the accumulated history. We propose to view KV-cache compression as a coreset selection problem: rather than scoring tokens independently for retention, we select a small subset that covers the geometry of the accumulated visual cache. Our method operates in a joint KV representation and introduces a bicriteria objective that balances coverage in key and value spaces, preserving both retrieval structure and output-relevant information. To encourage a more diverse retained subset, we further introduce an orthogonality-driven diversity criterion that favors candidates contributing new directions beyond the current selection, and connect this criterion to log-determinant subset selection. Across four open-source VLMs and five long-video and streaming-video benchmarks, our method improves over heuristic streaming compression baselines under a fixed cache budget. These results highlight that representative coreset selection offers a more effective principle, than token-wise pruning, for memory-constrained streaming video understanding.

Omnimodal large language models (Omni-LLMs) show strong capability in audio-video understanding, but their practical deployment remains limited by high inference cost of long video streams and dense audio sequences. Despite recent progress, existing compression methods for Omni-LLMs typically rely on fixed or native compression units, which can disrupt cross-modal correspondence and the complementary information required for audio-video reasoning, making it difficult to improve inference efficiency while stably preserving performance. To address this, we propose OmniRefine, a training-free two-stage framework for efficient audio-visual token compression in Omni-LLMs. First, Correspondence-Preserving Chunk Refinement refines native chunk boundaries into cross-modally aligned compression units through frame-audio similarity and dynamic programming. Second, Modality-Aware Cooperative Compression jointly compresses video and audio tokens within each refined unit to reduce redundancy while preserving critical evidence. Extensive experiments show that OmniRefine achieves a better efficiency-performance trade-off than strong baselines and maintains stable performance under lower compression ratios. On WorldSense, it still reaches 46.7% accuracy at a 44% token retention ratio, nearly matching the full-token baseline. The code and interface will be released to facilitate further research.

As Video Large Language Models (Video-LLMs) scale to longer and more complex videos, their inference cost grows rapidly due to the large volume of visual tokens accumulated across frames. Training-free token compression has emerged as a practical solution to this bottleneck. However, existing temporal compression methods rely primarily on cross-frame token similarity or segmentation heuristics, overlooking each token's semantic role within its frame and failing to adapt compression strength to the compressibility of each frame pair. In this work, we propose OTT-Vid, a transport-derived allocation framework for temporal token compression. Our approach consists of two stages: spatial pruning identifies representative content within each frame, and optimal transport (OT) is then solved between neighboring frames to estimate temporal compressibility. We formulate this OT with non-uniform token mass, which protects semantically important tokens from aggressive compression, and a locality-aware cost that captures both feature and spatial disparities. The resulting transport plan jointly balances token importance and matching cost, while its total cost defines the transport difficulty of each frame pair, which we use to allocate compression budgets dynamically. Experiments on six benchmarks spanning video question answering and temporal grounding show that OTT-Vid preserves 95.8% of VQA and 73.9% of VTG performance while retaining only 10% of tokens, consistently outperforming existing state-of-the-art training-free compression methods.

Semantic communication (SC) aims to reduce transmission overhead by conveying task-relevant information rather than raw data. However, existing SC approaches for video largely focus on pixel-level reconstruction or rely on complex spatiotemporal pipelines, leading to excessive bandwidth usage and latency that are unsuitable for low-resource deployments. In this paper, we propose ChronoSC, a task-oriented semantic communication framework for Video Question Answering (VideoQA). ChronoSC introduces Chrono-Color Stacking, a lightweight and lossless projection scheme that encodes temporal video dynamics into a single static image, enabling extreme temporal compression before transmission. This compact semantic representation is transmitted using a lightweight Deep Joint Source-Channel Coding (DeepJSCC) transceiver and explicitly reconstructed at the receiver. Unlike latent-space methods, explicit visual reconstruction enables the direct reuse of pre-trained vision-language models; specifically, a pre-trained BLIP model is employed to infer answers from noisy, reconstructed chrono-images. Experiments on the CLEVRER dataset show that ChronoSC achieves up to 192 times bandwidth reduction compared to raw video transmission while maintaining high VideoQA accuracy.

Accurate and effective discrete image tokenization is crucial for long image sequence processing. However, current methods rigidly compress all content at a fixed rate, ignoring the variable information density of images and leading to either redundancy or information loss. Inspired by information entropy, we propose TaTok, a Theoretically grounded adaptive image Tokenization framework. We rigorously identify two key drawbacks in existing methods: information insufficiency when reconstructing images with patch tokens alone, and information redundancy among patch tokens. To address these, we introduce global tokens that model mutual information across patch tokens, and a Dynamic Token Filtering (DTF) algorithm based on cumulative conditional entropy to eliminate redundancy. Experiments confirm TaTok's state-of-the-art performance, delivering a 1.3x gFID improvement and 8.7x inference speedup. By allocating tokens according to information richness, TaTok enables more compressed yet accurate image tokenization, offering valuable insights for future research.

Video MLLMs face a persistent tension between spatial fidelity and temporal coverage: preserving fine-grained visual details requires many spatial tokens, while capturing short-lived events requires dense temporal sampling. We propose \textbf{Fre-Res}, a budget-adaptive dual-track video-token compression framework that separates these two forms of evidence. Fre-Res preserves sparse high-fidelity spatial anchors and represents dense temporal evolution through compact residual-frequency tokens. Specifically, it applies temporal 1D-DCT to inter-frame residual trajectories in vision-latent space, where we observe strong low-frequency concentration. To align frequency-domain dynamics with native visual embeddings, Fre-Res introduces a Spatial-Guided Absorber that injects temporal residual information into spatially corresponding anchor tokens. Across fine-grained short-video and long-video reasoning benchmarks, Fre-Res achieves a favorable accuracy--efficiency trade-off, matching or approaching full-token performance while substantially reducing visual-token length. Extensive ablations further show that temporal-frequency residuals preserve causal transition cues, while spatial anchors remain essential for fine-grained object and layout reasoning.