The paper analyzes token reduction for efficient unified VLM training, finding that while task-specific acceleration saves computation, it destroys the mutual performance gains achieved through joint…

OccamToken introduces a training-free, adaptive token pruning framework that replaces fixed token budgets with relative evidence testing against a register-based reference, significantly improving VLM…

The paper proposes Cross-Layer Sparse Attention (CLSA) to significantly improve the efficiency and accuracy of long-context LLMs by jointly optimizing KV-cache sharing and the routing index across dec…

CIVIC is a path-consistent compact visual inference framework that achieves genuine hardware efficiency in Vision-Language Models by maintaining contiguous sequence representations across all inferenc…

The paper proposes Locality-Aware Redundancy Pruning (LoRP), a training-free method that prunes LLM layers by exploiting localized inter-layer redundancy, leading to improved efficiency while maintain…

PARCEL introduces a novel visual tokenization architecture that combines spatial pooling anchors with conditioned elastic queries, efficiently reducing the computational cost of large Vision-Language…

RASER introduces a family of cheap, router-based systems that selectively decide whether to perform expensive multi-hop retrieval, significantly reducing LLM token costs while maintaining state-of-the…

InfoMerge is a novel, training-free method that significantly compresses visual tokens for Video-LLMs by estimating temporal redundancy and allocating tokens based on content richness, achieving high…

Reasmory introduces a structured programming framework that uses explicit 3D memory and a Domain-Specific Language (DSL) to reliably enhance Vision-Language Models' spatial reasoning capabilities, ach…

V-LynX is a framework that enhances Video LLMs by integrating new modalities into their existing token interface, achieving state-of-the-art performance across diverse video understanding tasks.

The paper introduces D3IM, a novel parameter-free sampler that enables direct revision of visible tokens in Masked Diffusion Language Models, and proposes SCOPE to mitigate the model's tendency to per…

The paper proposes moving the query instead of the KV-cache during cross-instance attention, demonstrating that this approach is significantly cheaper than moving the cache, especially on modern GPU f…

AdaCodec introduces a predictive visual coding scheme for video MLLMs, significantly improving efficiency and performance by transmitting only inter-frame changes and full reference frames when necess…

PrunePath introduces a budget-adaptive structured sparsification framework that efficiently prunes Feed-forward networks in large language models, achieving hardware-friendly sparsity and measurable s…

ROVER is a lightweight, learnable plugin that efficiently routes and integrates object-centric visual evidence across multiple images and objects, significantly improving performance on grounded multi…

AlphaToken is a novel response token valuation framework that improves LLM post-training by decoupling token selection into task-specific adaptation and stability preservation, leading to better perfo…

The paper proposes Dynamic Adapter Routing (DAR), a novel method that significantly improves continual multimodal retrieval by adaptively selecting and merging specialized adapters.

The paper introduces LL-Bench, a comprehensive benchmark for evaluating large-scale generative models on low-level vision tasks, and proposes LL-Score, an MLLM-based evaluator that better aligns quali…