{"slug": "researchers-from-stanford-and-uc-berkeley-introduce-tokengate-token-gated-makes", "title": "Researchers from Stanford and UC Berkeley Introduce TokenGate: Token-Gated Attention Makes Long-Context Prefill Up to 11.3x Faster", "summary": "Researchers from Stanford and UC Berkeley released TokenGate, a drop-in learned token pruning module for large language models that cuts long-context prefill time-to-first-token by 1.4x to 11.3x and peak memory by 1.5x to 3.4x without accuracy loss. The module applies token-gated attention to prune tokens during prefill, targeting the cost of processing long inputs before generation begins.", "body_md": "Stanford and UC Berkeley researchers released TokenGate, a drop-in learned token pruning module for LLMs. It cuts long-context prefill time-to-first-token by 1.4x to 11.3x and peak memory by 1.5x to 3.4x without accuracy loss.\n\nThe post [Researchers from Stanford and UC Berkeley Introduce TokenGate: Token-Gated Attention Makes Long-Context Prefill Up to 11.3x Faster](https://www.marktechpost.com/2026/10/25/stanford-berkeley-tokengate-learned-token-pruning-llm-prefill/) appeared first on [MarkTechPost](https://www.marktechpost.com).", "url": "https://wpnews.pro/news/researchers-from-stanford-and-uc-berkeley-introduce-tokengate-token-gated-makes", "canonical_source": "https://www.marktechpost.com/2026/10/25/stanford-berkeley-tokengate-learned-token-pruning-llm-prefill/", "published_at": "2026-10-25 12:12:18+00:00", "updated_at": "2026-10-25 13:13:51.645047+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-research", "machine-learning"], "entities": ["Stanford University", "UC Berkeley", "TokenGate"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/researchers-from-stanford-and-uc-berkeley-introduce-tokengate-token-gated-makes", "markdown": "https://wpnews.pro/news/researchers-from-stanford-and-uc-berkeley-introduce-tokengate-token-gated-makes.md", "text": "https://wpnews.pro/news/researchers-from-stanford-and-uc-berkeley-introduce-tokengate-token-gated-makes.txt", "jsonld": "https://wpnews.pro/news/researchers-from-stanford-and-uc-berkeley-introduce-tokengate-token-gated-makes.jsonld"}}