Newest papers

TokenGate: Adapting LLMs for Fast and Memory-Efficient Prefill with Learned Token Pruning

Liana Patel, Rishabh Ranjan, Ion Gabriel Stoica, Matei Zaharia, Carlos Guestrin · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.26754 · cited by 0 works (OpenAlex) · open-access PDF

Arabic

وبدلاً من ذلك، فإننا نقوم بتطبيق TokenGate، وهو مكون صغير ومتحرك يزيد من المكونات التي تم تدريبها من قبل المختصين، لكي يتمكنوا من قراءة فقط ما يهمهم. على الجانب الآخر، فإن نموذج اللغات يقوم بتطبيق جميع تكنولوجيات الدخول من خلال كل طبقة من النماذج، مما يؤدي إلى تكاليف معالجة كبيرة للنسخة الطويلة. ونحن نقدم TokenGate، وهو مكون صغير ومتحرك يزيد من المكونات التي تم تدريبها من قبل المختصين، والذي يتيح تقسيم تكنولوجيات الدخول من نهاية إلى نهاية من أجل تسجيل الدخول من نهاية إلى نهاية من أجل تسجيل الدخول من نهاية إلى نهاية من أجل تسجيل الدخول من نهاية إلى نهاية من

Machine translation (Meta M2M100); the English original follows.

Abstract

Expert human readers process large inputs (e.g., code repositories and long documents) by skimming to closely read only what matters. On the other hand, language models process all input tokens through each model layer, resulting in significant processing costs for long-context inputs. We introduce TokenGate, a small, drop-in component that extends existing trained models to enable end-to-end learnable token pruning over long-context inputs for fast and memory-efficient prefill. To enable end-to-end learning despite the non-differentiable top-K operator for token pruning, we introduce token-gated attention, a mechanism inspired by Mixture-of-Experts that re-weights attention logits using token importance scores. Across diverse tasks, TokenGate achieves $1.4$-$11.3\times$ faster time-to-first token (TTFT) and $1.5$-$3.4\times$ lower peak memory at no accuracy degradation for prefill in transformers. Compared to the best-performing prior methods for token-pruning, TokenGate gives up to $6.4\times$ faster prefill and $2\times$ lower peak memory. TokenGate natively extends both standard transformers (e.g., Qwen3) and hybrid-attention models (e.g., Qwen3.5) with simple lightweight adaptation comparable to standard parameter-efficient fine-tuning. We make our code available at: https://github.com/tokengate-models/TokenGate.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Security token, Computer science, Pruning, Artificial intelligence, Process (computing), Language model

Is this paper real? This page was built from OpenAlex and arXiv records, not generated. Software can verify any reference in one call: GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.26754, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txt

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange