File size: 20,865 Bytes
52e97b8 dacdd5f 52e97b8 dacdd5f 52e97b8 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 | <!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1, viewport-fit=cover">
<meta name="claude:viewport-fit-sides" content="cover">
<title>CA3 · Language</title>
<meta name="description" content="Open language models learned 55 days of real news with CA3: better than RAG, at a seventh of the compute.">
<link rel="preconnect" href="https://fonts.googleapis.com">
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
<link rel="stylesheet" href="https://fonts.googleapis.com/css2?family=Source+Sans+3:wght@300;400;600;700&family=IBM+Plex+Mono:wght@400;500&display=swap">
<style>
:root{color-scheme:dark;--bg:#000;--bg2:#06080A;--panel:#0C1115;--line:rgba(255,255,255,.11);--line2:rgba(255,255,255,.06);
--ink:#F5F7F8;--ink2:#A4AFB5;--ink3:#7A868C;--blue:#7CC0E0;--red:#E8827C;--amber:#E9B36A;--green:#8FD6AE;
--f-ui:'Source Sans 3','Myriad Pro',Myriad,'Segoe UI',system-ui,-apple-system,Arial,sans-serif;
--f-mono:'IBM Plex Mono',ui-monospace,'SFMono-Regular',Menlo,Consolas,monospace;
--gl:max(20px,env(safe-area-inset-left,0px));--gr:max(20px,env(safe-area-inset-right,0px));
padding-top:env(safe-area-inset-top,0px);padding-bottom:env(safe-area-inset-bottom,0px)}
@media (min-width:980px){:root{--gl:max(40px,env(safe-area-inset-left,0px));--gr:max(40px,env(safe-area-inset-right,0px))}}
*{box-sizing:border-box}html{background:var(--bg);scroll-behavior:smooth}
body{margin:0;background:var(--bg);color:var(--ink);font-family:var(--f-ui);font-size:18px;line-height:1.5;font-variant-numeric:tabular-nums;-webkit-font-smoothing:antialiased}
a{color:var(--blue);text-decoration:none}a:focus-visible{outline:2px solid var(--blue);outline-offset:4px;border-radius:6px}
img,video{max-width:100%}h1,h2,h3{margin:0;text-wrap:balance}p{margin:0}[hidden]{display:none!important}
.wrap{max-width:1200px;margin:0 auto;padding-left:var(--gl);padding-right:var(--gr)}
header{position:sticky;top:0;z-index:20;background:rgba(0,0,0,.62);-webkit-backdrop-filter:blur(20px) saturate(180%);backdrop-filter:blur(20px) saturate(180%);border-bottom:1px solid var(--line2)}
header .wrap{display:flex;align-items:center;gap:22px;height:56px}
.brand{display:flex;align-items:center;gap:10px;font-weight:700;font-size:19px;letter-spacing:.08em;color:var(--ink)}.brand img{width:24px;height:24px}
.back{margin-left:auto;font-size:15px;color:var(--ink2)}.back:hover{color:var(--ink)}
.pill{font-size:14px;font-weight:600;color:#000;background:var(--ink);border-radius:999px;padding:7px 14px;white-space:nowrap}
.eyebrow{font-size:14px;letter-spacing:.22em;text-transform:uppercase;color:var(--blue);font-weight:600}
.hero{padding-block:clamp(56px,9vh,110px) 40px;background:radial-gradient(70% 60% at 50% 30%,#0F1C24 0%,#05090C 55%,#000 100%)}
.hero .head{display:flex;flex-direction:column;gap:20px;align-items:center;text-align:center}
.hero h1{font-size:clamp(38px,5.4vw,76px);line-height:1.03;font-weight:600;letter-spacing:-.03em;max-width:20ch}
.hero .sub{font-size:clamp(19px,1.8vw,23px);color:var(--ink2);max-width:46ch;line-height:1.42}
.hero .media{margin-top:44px;border-radius:20px;overflow:hidden;border:1px solid var(--line);background:#05080A}
.hero .media img,.hero .media video{display:block;width:100%;max-height:78vh;object-fit:contain}
.demo{display:inline-block;font-size:11px;letter-spacing:.14em;text-transform:uppercase;color:var(--ink3);border:1px solid var(--line);border-radius:6px;padding:3px 8px}
.specs{display:grid;grid-template-columns:repeat(4,minmax(0,1fr));margin-top:48px;border-top:1px solid var(--line)}
.specs div{padding:22px 18px 0 0;display:flex;flex-direction:column;gap:6px;min-width:0}
.specs .k{font-size:12px;letter-spacing:.2em;text-transform:uppercase;color:var(--ink3);font-weight:600}
.specs .v{font-size:clamp(28px,3vw,46px);font-weight:300;line-height:1.05;letter-spacing:-.03em;color:var(--blue)}
.specs .c{font-size:15px;color:var(--ink2)}
section.block{padding-block:clamp(70px,10vh,120px);border-top:1px solid var(--line2)}
section.block.alt{background:var(--bg2)}
.bh{display:flex;flex-direction:column;gap:14px;max-width:760px}
.bh .verb{font-size:13px;letter-spacing:.22em;text-transform:uppercase;font-weight:600;color:var(--blue)}
.bh .verb.r{color:var(--red)}.bh .verb.g{color:var(--ink3)}
.bh h2{font-size:clamp(30px,3.6vw,52px);font-weight:600;letter-spacing:-.025em;line-height:1.06}
.bh p{color:var(--ink2);font-size:clamp(18px,1.6vw,21px);line-height:1.45}
.pair{display:grid;grid-template-columns:repeat(2,minmax(0,1fr));gap:8px;margin-top:40px}
.pair.three{grid-template-columns:repeat(3,minmax(0,1fr))}
.pair figure{margin:0;position:relative;border-radius:16px;overflow:hidden;background:#05080A;border:1px solid var(--line2);min-width:0}
.pair img,.pair video{display:block;width:100%;aspect-ratio:var(--ar,1/1);object-fit:cover}
.pair .px img{image-rendering:pixelated}
.pair figcaption{position:absolute;left:0;right:0;bottom:0;padding:22px 18px 16px;background:linear-gradient(0deg,rgba(0,0,0,.85),rgba(0,0,0,0));display:flex;flex-direction:column;gap:2px}
.pair .who{font-size:12px;letter-spacing:.18em;text-transform:uppercase;font-weight:600;color:var(--ink)}
.pair .num{font-size:clamp(34px,3.8vw,56px);font-weight:300;line-height:1;letter-spacing:-.035em}
.pair .bad .num{color:var(--red)}.pair .good .num{color:var(--blue)}.pair .neutral .num{color:var(--ink)}
.pair .what{font-size:14px;color:var(--ink2)}
.kp{display:flex;gap:40px;margin-top:30px;flex-wrap:wrap}
.kp b{display:block;font-size:clamp(36px,3.8vw,56px);font-weight:300;line-height:1;letter-spacing:-.035em;color:var(--blue);white-space:nowrap}
.kp b.r{color:var(--red)}.kp b.w{color:var(--ink)}
.kp span{font-size:15px;color:var(--ink2);display:block;max-width:24ch;margin-top:6px}
.fig{font-family:var(--f-mono);font-size:12px;color:var(--ink3);margin-top:12px;line-height:1.6}
.chips{display:flex;gap:8px;flex-wrap:wrap;margin-top:22px}
.chips span{font-family:var(--f-mono);font-size:12.5px;border:1px solid var(--line);border-radius:999px;padding:4px 10px;color:var(--ink2)}
.quote{margin-top:26px;font-family:var(--f-mono);font-size:clamp(15px,1.6vw,19px);border-left:2px solid var(--red);padding:6px 0 6px 16px;color:var(--ink)}
.quote small{display:block;color:var(--ink3);font-size:13px;margin-top:6px}
.quote.b{border-left-color:var(--blue)}.quote.g{border-left-color:var(--ink3)}
.charts{display:grid;grid-template-columns:repeat(2,minmax(0,1fr));gap:14px;margin-top:40px}
.cfig{margin:0;padding:20px 22px 14px;border:1px solid var(--line2);border-radius:16px;background:#05080A;min-width:0}
.cfig figcaption{font-size:16px;color:var(--ink);margin-bottom:12px}.cfig figcaption span{color:var(--ink3);font-size:13px;margin-left:6px}
.cfig svg{display:block;width:100%;height:auto;overflow:visible}
.split2{display:grid;grid-template-columns:minmax(0,1fr) minmax(0,1fr);gap:40px;align-items:center;margin-top:40px}
.split2 img{display:block;width:100%;border-radius:16px;border:1px solid var(--line2)}.split2 .kp{margin-top:0;flex-direction:column;gap:26px}
.fp{margin-top:40px}.fp table{min-width:0}.fp td:nth-child(2){font-family:var(--f-mono);color:var(--blue)}
.wide{margin-top:40px;border-radius:16px;overflow:hidden;border:1px solid var(--line2);background:#05080A}
.wide img,.wide video{display:block;width:100%}
.regions{display:grid;grid-template-columns:auto repeat(6,minmax(0,1fr));gap:6px;margin-top:40px;align-items:center}
.regions .rl{font-size:12px;letter-spacing:.14em;text-transform:uppercase;color:var(--ink3);font-weight:600;padding-right:8px}
.regions .ch{font-size:14px;font-weight:600;text-align:center}.regions .ch small{display:block;font-weight:400;color:var(--ink3);font-size:12px}
.regions img{display:block;width:100%;border-radius:8px}
.regions .acc{text-align:center;font-size:clamp(15px,1.5vw,20px);font-weight:300}.regions .acc b{color:var(--blue);font-weight:400}
.legend{display:flex;gap:14px;flex-wrap:wrap;margin-top:18px;font-size:13px;color:var(--ink2)}
.legend i{display:inline-block;width:10px;height:10px;border-radius:2px;margin-right:6px;vertical-align:-1px}
.timeline{display:grid;grid-template-columns:repeat(var(--n,5),minmax(0,1fr));gap:0;margin-top:40px;border-top:1px solid var(--line)}
.timeline div{padding:18px 16px 0 0;min-width:0}
.timeline .w{font-family:var(--f-mono);font-size:12px;color:var(--ink3)}
.timeline b{display:block;font-size:17px;font-weight:600;margin:6px 0 4px}
.timeline span{font-size:15px;color:var(--ink2)}
.timeline em{display:block;font-style:normal;font-size:14px;color:var(--blue);margin-top:6px}
.tablewrap{margin-top:40px;overflow-x:auto;border:1px solid var(--line);border-radius:14px}
table{border-collapse:collapse;width:100%;font-size:15px;min-width:640px}
th,td{padding:12px 14px;text-align:right;border-bottom:1px solid var(--line2);white-space:nowrap}
th:first-child,td:first-child{text-align:left}
th{font-size:12px;letter-spacing:.12em;text-transform:uppercase;color:var(--ink3);font-weight:600}
tr.ca3 td{color:var(--blue)}tr.ca3 td:first-child{font-weight:600}
tr:last-child td{border-bottom:0}
.takeaways{display:grid;grid-template-columns:repeat(3,minmax(0,1fr));gap:26px;margin-top:40px}
.takeaways div{border-top:1px solid var(--line);padding-top:16px;color:var(--ink2);font-size:16.5px}
.film{margin-top:40px}.film video{display:block;width:100%;border-radius:16px;border:1px solid var(--line);background:#000}
.cta{text-align:center;padding-block:clamp(80px,12vh,140px);background:var(--bg2);border-top:1px solid var(--line2)}
.cta h2{font-size:clamp(34px,4.6vw,64px);font-weight:600;letter-spacing:-.028em}
.cta p{color:var(--ink2);font-size:clamp(18px,1.6vw,21px);max-width:44ch;margin:18px auto 0}
.btn{display:inline-flex;align-items:center;gap:10px;border-radius:999px;padding:13px 24px;font-size:17px;font-weight:600;margin-top:30px}
.btn.pri{background:var(--ink);color:#000}.btn.ghost{border:1px solid rgba(255,255,255,.3);color:var(--ink);margin-left:10px}
.notes{border-top:1px solid var(--line2)}.notes .wrap{padding-block:50px 56px;display:grid;grid-template-columns:repeat(3,minmax(0,1fr));gap:34px}
.notes h3{font-size:13px;letter-spacing:.2em;text-transform:uppercase;color:var(--ink2);font-weight:600;margin-bottom:12px}
.notes p,.notes li{font-size:13px;color:var(--ink3);line-height:1.6}.notes ul{margin:0;padding-left:18px}.notes .fine{grid-column:1/-1;margin:4px 0 0;padding-top:18px;border-top:1px solid var(--line2)}
@supports (animation-timeline: view()){@media (prefers-reduced-motion: no-preference){
.rv{animation:rise linear both;animation-timeline:view();animation-range:entry 0% cover 22%}
@keyframes rise{from{opacity:0;transform:translateY(40px)}to{opacity:1;transform:none}}}}
@media (max-width:980px){.specs{grid-template-columns:repeat(2,minmax(0,1fr))}.specs div{padding-bottom:20px}
.takeaways,.notes .wrap{grid-template-columns:minmax(0,1fr)}.timeline{grid-template-columns:repeat(2,minmax(0,1fr))}.timeline div{padding-bottom:18px}
.regions{grid-template-columns:repeat(3,minmax(0,1fr))}.regions .rl{grid-column:1/-1;padding-top:10px}.back{display:none}.pill{margin-left:auto}}
@media (max-width:620px){.pair,.pair.three,.charts,.split2{grid-template-columns:minmax(0,1fr)}.timeline{grid-template-columns:minmax(0,1fr)}
.regions{grid-template-columns:repeat(2,minmax(0,1fr))}.btn.ghost{margin-left:0}}
.when{font-size:14px;letter-spacing:.04em;color:var(--ink3)}</style>
<link rel="canonical" href="https://ca3.si/case_language.html">
</head>
<body>
<header><div class="wrap"><a class="brand" href="index.html"><img src="ammonite_dark.svg" alt="">CA3</a>
<a class="back" href="index.html#cases">All use cases</a><a class="pill" href="index.html#pilot">Contact</a></div></header>
<section class="hero"><div class="wrap"><div class="head"><span class="eyebrow">Language · Qwen · Llama · SmolLM2 · NVIDIA Nemotron</span><h1>It learns the news every day. More right than RAG, at a seventh of the compute.</h1><p class="sub">In a 55-day test on real news from 2020, CA3 let open language models learn every day and answer from what they learned, with no lookup. What it learns is checked first; the original model never changes.</p><span class="when">Results recorded on 21 October 2026</span></div><div class="media rv"><img src="lang_overview.jpg" alt="The CA3 console on day 40 of the recorded test" loading="lazy"></div><div class="specs rv"><div><span class="k">Compute per answer</span><span class="v">7×</span><span class="c">less than RAG; 4.5× on Nemotron 3 Nano</span></div><div><span class="k">Wrong answers</span><span class="v">−25%</span><span class="c">against RAG: 6 against 8 of 86</span></div><div><span class="k">Planted falsehoods</span><span class="v">0 of 24</span><span class="c">kept; a RAG repeated 8</span></div><div><span class="k">Erased facts</span><span class="v">0</span><span class="c">said again (0 of 6, 0 of 7)</span></div></div></div></section><section class="block"><div class="wrap"><div class="bh rv"><span class="verb ">Learns</span><h2>The same fact, asked with no documents.</h2><p>The original model gets it wrong. After reading the news, CA3 answers from long-term memory, with 17 times less compute than reading the news again.</p></div><div class="wide rv"><img src="lang_compare.jpg" alt="The original model answers Baanke Thakur, wrong; CA3 on day 19 answers Mirza, right" loading="lazy"></div><div class="kp rv"><div><b class="">93%</b><span>right on day 55, asked in words it never saw; RAG 91% (Alibaba Qwen3 14B)</span></div><div><b class="">≤0.005</b><span>general knowledge lost in 55 days (MMLU, NaturalQuestions, TriviaQA)</span></div></div></div></section><section class="block alt"><div class="wrap"><div class="bh rv"><span class="verb r">Blocks</span><h2>24 credible falsehoods were planted in the news.</h2><p>What contradicts what it learned is held back, and the doubtful cases go to a person with the passages side by side. CA3 kept none; a RAG over the same news repeated 8.</p></div><div class="wide rv"><img src="lang_review.jpg" alt="The review queue: a planted falsehood shown with its passages" loading="lazy"></div><div class="kp rv"><div><b class="">0 of 24</b><span>planted falsehoods kept (NVIDIA Nemotron 3 Nano 30B)</span></div><div><b class="r">8</b><span>repeated by a RAG over the same news</span></div><div><b class="r">7</b><span>kept by the same learning without CA3's control</span></div></div></div></section><section class="block"><div class="wrap"><div class="bh rv"><span class="verb g">Forgets</span><h2>Ten items erased in 108 seconds.</h2><p>On day 40 an erasure request arrives. CA3 removes what it learned from them, checks again, and everything else stays as right as before.</p></div><div class="wide rv"><img src="lang_deletions.jpg" alt="The erasure job in the console" loading="lazy"></div><div class="kp rv"><div><b class="">108 s</b><span>to erase ten items</span></div><div><b class="">99.1%</b><span>everything else, right before and after</span></div><div><b class="w">1 of 10</b><span>still answered: the original model already knew it</span></div></div></div></section><section class="block alt"><div class="wrap"><div class="bh rv"><span class="verb ">Replay</span><h2>The 55 days, in 90 seconds.</h2><p>A replay of the recorded test, and a console to walk through it day by day: what it read, where each answer came from, each case sent to the reviewer, the erasure and the audit log.</p></div><div class="film rv"><video controls muted playsinline preload="none" poster="ca3_replay_poster.jpg"><source src="ca3_replay_b.mp4" type="video/mp4"></video></div><p class="fig">Replay of a recorded, pre-registered test · reviewer simulated</p><a class="btn pri" href="replay.html">Open the replay console</a></div></section><section class="block"><div class="wrap"><div class="bh rv"><span class="verb ">Round by round</span><h2>Earlier rounds, smaller models.</h2><p>The best figure measured in each comparison, from rounds on models from 0.5B to 8B.</p></div><div class="kp rv"><div><b class="">0–6%</b><span>contradictions of what it learned, adopted; 37–56% without CA3 (Qwen2.5 7B, Llama 3.1 8B)</span></div><div><b class="">0.80–0.87</b><span>useful accuracy after 40 sessions; ordinary fine-tuning 0.52–0.53</span></div><div><b class="">40%</b><span>of the learning time to carry what it learned to a new base model (Qwen2.5 7B to Qwen3 8B Base)</span></div><div><b class="">$2–4</b><span>of GPU a month (H200) for one customer's daily learning, in the 55-day test</span></div></div></div></section><section class="block alt"><div class="wrap"><div class="bh rv"><span class="verb ">How the test was run</span><h2>The test, step by step.</h2><p>A simulated company on public news: 40 days of news as it was published, 15 more days of questions, with planted falsehoods and an erasure request.</p></div><div class="timeline rv" style="--n:4"><div><span class="w">Days 1 to 40</span><b>The news arrives</b><span>11,642 passages from WMT News Crawl 2020, read as they were published.</span><em>Updates every day; 1 of 29 rolled back</em></div><div><span class="w">Along the way</span><b>24 planted falsehoods</b><span>Credible falsehoods, planted in the news.</span><em>0 of 24 kept</em></div><div><span class="w">Day 40</span><b>An erasure request</b><span>Ten items to remove.</span><em>Done in 108 s</em></div><div><span class="w">Day 55</span><b>The final test</b><span>Questions about what it learned, asked again in other words, against a RAG over the same news.</span><em>6 of 6 tests passed</em></div></div><p class="fig">Pre-registered round 20d, on Alibaba Qwen3 14B and NVIDIA Nemotron 3 Nano 30B, one run each on one H200. The reviewer was simulated.</p><div class="tablewrap rv"><table><thead><tr><th>Day 55</th><th>Right, CA3</th><th>Right, RAG</th><th>Compute per answer</th></tr></thead><tbody><tr class="ca3"><td>Alibaba Qwen3 14B (86 questions)</td><td>93.0%</td><td>90.7%</td><td>7× less</td></tr><tr class="ca3"><td>NVIDIA Nemotron 3 Nano 30B (96 questions)</td><td>86.5%</td><td>83.3%</td><td>4.5× less</td></tr></tbody></table></div><div class="takeaways rv"><div>On both models CA3 answered more questions right than the RAG, with less compute.</div><div>85% of Nemotron's answers came from long-term memory, with no lookup.</div><div>93 in 100 checks were settled on its own; 229 of 3,136 went to a person, on three models.</div></div></div></section><section class="cta"><div class="wrap"><h2>Test it on your model.</h2><p>Four to eight weeks on your own data and your own model: accuracy and cost against your RAG, what it learned, what it refused and what it forgot.</p><a class="btn pri" href="index.html#pilot">Contact</a><a class="btn ghost" href="index.html#cases">All use cases</a></div></section><section class="notes"><div class="wrap"><div><h3>Model and data</h3><ul><li>Alibaba Qwen2.5 0.5B, 1.5B and 7B, Qwen3 8B Base and 14B; Meta Llama 3.1 8B; Hugging Face SmolLM2 1.7B; NVIDIA Nemotron 3 Nano 30B (mixture of experts).</li><li>News: WMT News Crawl 2020, with StreamingQA questions.</li><li>Every figure comes from a pre-registered round and its reading; “right” is against the answer key of the test.</li></ul></div><div><h3>Sources</h3><ul><li>Liška A, et al. StreamingQA: a benchmark for adaptation to new knowledge over time in question answering models. ICML 2022.</li></ul></div><div><h3>Limits</h3><ul><li>The 7× is on the day-55 test questions; on the traffic of days 31 to 40, answering cost 1.8 to 2.2 times less than the RAG.</li><li>Counting the daily learning, the whole cycle costs less than the RAG above about 140 questions a day (Qwen3 14B, computed from the measured figures).</li><li>Falsehoods about facts the model already knew before are adopted about as often as with RAG.</li><li>One simulated company on public news, one run per model; the 14B's two planted falsehoods on day 40 came with the documents in front of it.</li></ul></div><p class="fine">Results from recorded, pre-registered tests. The mechanism is not shown. Model and dataset names are for identification only. Not produced or endorsed by Alibaba Cloud, NVIDIA, Meta or Hugging Face. CA3 is developed by Alberto Acedo, PhD, at Biome Makers Inc.</p></div></section>
<script>
(function(){var still=window.matchMedia&&window.matchMedia('(prefers-reduced-motion: reduce)').matches;
var vids=[].slice.call(document.querySelectorAll('video.loop'));
if(still||!('IntersectionObserver' in window)){vids.forEach(function(v){v.controls=true;});return;}
var io=new IntersectionObserver(function(es){es.forEach(function(e){var v=e.target;if(e.isIntersecting){var p=v.play();if(p&&p.catch)p.catch(function(){v.controls=true;});}else v.pause();});},{threshold:0.3});
vids.forEach(function(v){io.observe(v);});})();
</script>
</body>
</html>
|