[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"branding":3,"analytics":7,"article-freqdepthkv-cuts-llm-memory-use-without-wrecking-accuracy":10,"sections":40},{"siteName":4,"siteTagline":5,"publisherName":4,"contactEmail":6},"The Revision","Tech news, decoded.","editor@therevision.news",{"gaMeasurementId":8,"adsenseClientId":9},"G-ZW2MV82GYR","ca-pub-8533917693782264",{"article":11},{"id":12,"slug":13,"title":14,"dek":15,"body_md":16,"tags_json":17,"published_at":18,"created_at":19,"updated_at":20,"status":21,"review_note":22,"review_notes":23,"image_url":22,"persona_id":22,"persona_name":22,"section":30,"tags":31,"sources":35,"feedback":39,"feedback_at":22,"cost_usd":39,"total_tokens":39},4305,"freqdepthkv-cuts-llm-memory-use-without-wrecking-accuracy","FreqDepthKV Cuts LLM Memory Use Without Wrecking Accuracy","A new inference-time method from arXiv:2607.06519 compresses key-value caches 3.9x while preserving benchmark scores on long-context tasks.","A research paper proposes a smarter way to shrink the memory footprint of large language models during inference — without the usual accuracy penalty.\n\nThe method, called FreqDepthKV, targets the key-value cache: the working memory an LLM builds up as it processes a long prompt. As context windows stretch to tens of thousands of tokens, that cache becomes a bottleneck — both for RAM and for how fast tokens can be generated. The researchers behind arXiv:2607.06519 split adjacent-layer cache states into shared low-frequency components (structure that repeats across layers) and sparse high-frequency residuals (the layer-specific detail). A lightweight online probe then decides, per attention head, which mode to use — shared, residual, or exact — based on how sensitive that head is to reconstruction error. No retraining required.\n\nThe numbers hold up on a 32k-token prefill: 58.3 Exact Match and 63.0 F1 on question answering, 32.5 ROUGE-L on summarization, and 48.1 pass@1 on code generation — all close to full-cache baselines while beating prior compression methods. Peak KV memory drops to 6.2 GB, decoding throughput rises to 70.4 tokens\u002Fs, and time-to-first-token falls to 2.06 seconds, at a 3.9x compression ratio.\n\nKV cache compression has become a crowded research space — methods like H2O, SnapKV, and MagicPLD all chip away at the same problem — but most require either retraining or a fixed compression policy that ignores prompt structure. The per-head, prompt-adaptive framing here is the actual contribution worth watching, assuming it survives scrutiny beyond the benchmarks reported.","[\"ai\",\"llm\",\"inference\",\"research\"]","2026-07-08T04:00:00.000Z","2026-07-08T05:15:10.034Z","2026-07-08T05:15:12.851Z","published",null,[24],{"id":25,"reviewer":26,"round":27,"reason":28,"status":29},"editor-r1","editor",1,"The article omits author attribution and does not identify the institution or paper source (arXiv:2607.06519), which are required for the piece to be independently verifiable.","resolved","ai",[30,32,33,34],"llm","inference","research",[36],{"name":37,"url":38},"arXiv cs.AI","https:\u002F\u002Farxiv.org\u002Fabs\u002F2607.06519",0,{"sections":41},[42,46,51,56,61,66,71,76,81,86,91,95,100,105],{"name":43,"slug":30,"count":44,"latest_published_at":45},"AI",2590,"2026-07-16T04:00:00.000Z",{"name":47,"slug":48,"count":49,"latest_published_at":50},"Security","security",294,"2026-07-15T19:59:48.000Z",{"name":52,"slug":53,"count":54,"latest_published_at":55},"Deals","deals",179,"2026-06-29T20:02:07.000Z",{"name":57,"slug":58,"count":59,"latest_published_at":60},"Policy","policy",158,"2026-07-16T00:02:48.000Z",{"name":62,"slug":63,"count":64,"latest_published_at":65},"Hardware","hardware",122,"2026-07-14T19:46:26.000Z",{"name":67,"slug":68,"count":69,"latest_published_at":70},"Consumer Tech","consumer-tech",93,"2026-07-13T13:20:48.000Z",{"name":72,"slug":73,"count":74,"latest_published_at":75},"Software","software",70,"2026-07-13T19:52:25.000Z",{"name":77,"slug":78,"count":79,"latest_published_at":80},"Science","science",66,"2026-07-10T10:29:37.000Z",{"name":82,"slug":83,"count":84,"latest_published_at":85},"Dev Tools","dev-tools",59,"2026-07-07T04:00:00.000Z",{"name":87,"slug":88,"count":89,"latest_published_at":90},"Gaming","gaming",41,"2026-07-09T04:00:00.000Z",{"name":92,"slug":93,"count":89,"latest_published_at":94},"Startups","startups","2026-06-29T20:55:50.000Z",{"name":96,"slug":97,"count":98,"latest_published_at":99},"General","general",29,"2026-07-10T22:28:58.000Z",{"name":101,"slug":102,"count":103,"latest_published_at":104},"Reviews","reviews",20,"2026-06-24T12:00:01.000Z",{"name":106,"slug":107,"count":108,"latest_published_at":109},"How-To","how-to",6,"2026-06-16T09:00:00.000Z"]