SpecFlow is proposed, a training-free framework that shifts the paradigm from destructive pruning to conservative condensation, strictly enforcing spatial coverage and statistical conservation to ensure stability.
Abstract
Vision-Language Models (VLMs) are costly at inference time because they must process long sequences of visual tokens. Existing token pruning methods often degrade under high compression by blindly discarding information, breaking spatial structure or collapsing diversity. We propose SpecFlow, a training-free framework that shifts the paradigm from destructive pruning to conservative condensation, strictly enforcing spatial coverage and statistical conservation to ensure stability. Treating visual tokens as nodes in a $k$NN graph, SpecFlow (i) computes a stable importance field via spectral heat flow to preserve structural coherence, (ii) allocates budgets via adaptive spatial partitioning to guarantee coverage, and (iii) aggregates discarded information into coreset sinks to maintain statistical conservation. The method is plug-and-play, requires no fine-tuning, and is compatible with FlashAttention. Experiments confirm that our SpecFlow outperforms SOTA methods across tasks, VLM architectures, and pruning ratios. Notably, LLaVA-1.5 with SpecFlow retains 95.6% of original performance despite pruning 88.9% of visual tokens, offering an exceptional efficiency-accuracy balance. Code is available at https://github.com/Lzy-dot/SpecFlow
While visual token pruning is essential for efficient Multimodal Large Language Models (MLLMs), existing training-free methods suffer from a critical limitation: they rely on static, instantaneous heuristics to perform irreversible filtering. This approach ignores the hierarchical nature of MLLMs, where token importance often evolves dynamically rather than remaining fixed across layers. Consequently, tokens essential for deep-layer reasoning are often prematurely discarded by shallow-layer estimates. To address this, we propose Trend-aware Pruning, a novel framework that elevates pruning from a local snapshot decision to a temporal trajectory modeling problem. Instead of relying on isolated scores, our method captures the momentum of attention flow. This enables a dynamic rectification mechanism that selectively reactivates"late-blooming"tokens, those initially undervalued but exhibiting rising semantic importance, thereby preventing the loss of critical visual cues. Extensive experiments demonstrate that our approach achieves a superior efficiency-performance trade-off across diverse multimodal tasks. Notably, it reduces visual tokens by over 77.8%, retaining only approximately 23 tokens in the final layer while maintaining competitive performance, offering a robust and reversible solution for high-efficiency multimodal inference.
Recent high-resolution Multimodal Large Language Models (MLLMs) generate thousands of visual tokens per input, leading to a visual token explosion that introduces severe latency bottlenecks. While token pruning mitigates this issue, state-of-the-art subset-optimization methods typically rely on iterative subset construction to jointly capture visual diversity and instruction relevance. As visual token counts scale, this sequential dependency introduces significant selection overhead, severely limiting the translation of theoretical FLOPs reductions into actual wall-clock speedups. To address this limitation, we propose Single-Forward Pruner (SFPruner), a structural reformulation of visual token pruning that embeds redundancy control directly into the scoring space, bypassing the need for iterative combinatorial optimization. Our non-iterative framework achieves redundancy-aware importance selection in a single forward pass through two complementary mechanisms. First, to attenuate redundancy at the covariance level, we introduce a semantics-guided ridge leverage scheme. By integrating instruction relevance and visual saliency, this mechanism suppresses dominant covariance directions and mitigates representation bias. Second, ranking-based directional masking resolves residual overlap through asymmetric similarity competition, where higher-scoring tokens explicitly suppress redundant lower-scoring alternatives via parallel tensor operations. Extensive evaluations demonstrate that our approach maintains stable selection costs, reducing the token selection process by up to 110 ms, from 112.4 ms to just 2.5 ms at 512 tokens in Qwen2.5-VL. This structural efficiency successfully translates theoretical token reductions into tangible inference speedups while preserving highly competitive performance against state-of-the-art techniques under aggressive compression.
J. Song, Woohyeong Kim, Kyeongbo Kong· 0 citations
While Vision Large Language Models (VLLMs) have achieved remarkable success in multimodal reasoning, their long-context inference remains prohibitively expensive due to the massive computation and memory overhead of visual Key-Value (KV) caches. Existing KV compression methods often apply uniform pruning across visual tokens and layers, leading to substantial information loss and degraded performance.To address this challenge, we propose \textbf{VisCache}, a plug-and-play framework for coarse-to-fine \textbf{Vis}ual KV \textbf{Cache} pruning without training, which consists of two synergistic stages. First, a lightweight VLM filters temporal redundancy by selectively forwarding semantically informative keyframes. Second, we introduce {PruneKV}, a surgical KV compression algorithm tailored to the attention dynamics of VLLMs. Unlike rigid pruning strategies, PruneKV adopts a parabolic layer-wise budget allocation together with an asymmetric update mechanism that selectively prunes keys while fusing values, thereby preserving critical contextual information. Extensive experiments demonstrate that VisCache substantially improves inference efficiency, achieving up to {2.35$\times$ speedup} and significant memory reduction while maintaining competitive performance with only {19--28\%} KV cache retention. VisCache consistently outperforms existing baselines, establishing a new Pareto frontier between efficiency and performance for long-context VLLM inference. Code is available at https://github.com/Wlklk/VisCache
Vision-Language Models (VLMs) have achieved strong performance in multimodal understanding, yet remain challenging to deploy on resource-constrained edge devices due to the substantial computational overhead of processing numerous visual tokens. Token reduction is a promising direction for accelerating VLMs inference, but existing approaches either rely on attention maps that are incompatible with modern acceleration frameworks or depend on computationally intensive pairwise similarity comparisons, which undermine scalability and negate their practical benefits in deployment. In this paper, we propose an attention-free and lightweight token reduction framework as a plug-and-play module for VLMs, which preserves both important and diverse tokens to produce a compact visual representation. First, to enable attention-free importance estimation, we adopt an information-theoretic perspective and quantify token information using a novel entropy-based criterion, retaining those with more expressive and less degenerate feature representations. Second, to ensure diverse visual coverage in a lightweight manner, we introduce a transformation-induced consistency signal where similar tokens yield similar signals, such that sorting by this signal places similar tokens close to each other and enables stride-based selection to produce a diverse token set. Extensive experiments across multiple VLMs benchmarks demonstrate that our framework achieves a favorable accuracy-efficiency trade-off, maintaining competitive performance under aggressive compression.
Xuanyi Hao, Zuoyuan Zhang, Zhibo Wang et al.· 0 citations
Long-video understanding is bottlenecked by the high cost of processing massive visual to-kens. Current reduction strategies often rely on static allocation or inefficient in-network selection that disrupts optimized attention kernels. In this paper, we introduce Vista-LLM, a decoupled framework for query-guided visual token pruning. By filtering redundancy prior to inference with minimal overhead, Vista-LLM ensures full compatibility with Flash Attention. Our method employs a coarse-to-fine pipeline: (1) Query-Guided Dynamic Bud-geting for adaptive temporal allocation; (2) a lightweight Semantic Scout for fine-grained, query-specific selection; and (3) Structure-Aware Compensation to preserve global context. Extensive experiments on benchmarks like Video-MME and MLVU demonstrate a significantly improved Pareto frontier. No-tably, on LLaVA-OneVision, Vista-LLM reduces visual tokens by 90% and accelerates inference while retaining over 98% of baseline performance on average, effectively filtering visual noise. Our code is available at https: //github.com/lizhenyu-123/Vista-LLM .
Zhenyu Li, Zuchao Li, Ping Wang et al.· Annual Meeting of the Associ...· 0 citations
Visual token pruning reduces the computational cost of Vision-Language Models (VLMs) by removing redundant visual tokens. The key is to learn a score that measures whether a token is useful. Existing methods typically rely on Gumbel-Softmax to approximate discrete selection during training. Such selectors make the score depend on the behavior of a relaxed pruning operator, not directly on the consequence of information loss. In this paper, we propose DiffPrune, which gives token scores a direct meaning. During training, DiffPrune keeps all tokens and weakens each token's information according to its score. If weakening a token hurts the task, the scorer is pushed to protect it; if not, the token can receive a lower score. Because the loss is differentiated through this actual information-throttling path, the scorer avoids the unstable surrogate path of relaxed token selection. DiffPrune implements this idea with an Information Throttler, which injects variance-preserving noise into visual tokens, where high-score tokens remain close to their original representations, while low-score tokens carry less original information. At inference, the throttler is removed, and hard top-K pruning is applied using the learned scores. Across ten VLM benchmarks, DiffPrune retains 96.5% of full-model accuracy while accelerating LLM prefill by 2.85x, with only 0.69 ms inference overhead. Code will be publicly available.