{"version":"1.0","provider_name":"Megagon","provider_url":"https:\/\/megagon.ai\/jp\/","author_name":"createdbyred-team","author_url":"https:\/\/megagon.ai\/jp\/author\/createdbyred-team\/","title":"Holistic Reasoning with Long-Context LMs: A Benchmark for Database Operations on Massive Textual Data","type":"rich","width":600,"height":338,"html":"<blockquote class=\"wp-embedded-content\" data-secret=\"XltIPWarUC\"><a href=\"https:\/\/megagon.ai\/jp\/publications\/holistic-reasoning-with-long-context-lms\/\">Holistic Reasoning with Long-Context LMs: A Benchmark for Database Operations on Massive Textual Data<\/a><\/blockquote><iframe sandbox=\"allow-scripts\" security=\"restricted\" src=\"https:\/\/megagon.ai\/jp\/publications\/holistic-reasoning-with-long-context-lms\/embed\/#?secret=XltIPWarUC\" width=\"600\" height=\"338\" title=\"&#8220;Holistic Reasoning with Long-Context LMs: A Benchmark for Database Operations on Massive Textual Data&#8221; &#8212; Megagon\" data-secret=\"XltIPWarUC\" frameborder=\"0\" marginwidth=\"0\" marginheight=\"0\" scrolling=\"no\" class=\"wp-embedded-content\"><\/iframe><script>\n\/*! This file is auto-generated *\/\n!function(d,l){\"use strict\";l.querySelector&&d.addEventListener&&\"undefined\"!=typeof URL&&(d.wp=d.wp||{},d.wp.receiveEmbedMessage||(d.wp.receiveEmbedMessage=function(e){var t=e.data;if((t||t.secret||t.message||t.value)&&!\/[^a-zA-Z0-9]\/.test(t.secret)){for(var s,r,n,a=l.querySelectorAll('iframe[data-secret=\"'+t.secret+'\"]'),o=l.querySelectorAll('blockquote[data-secret=\"'+t.secret+'\"]'),c=new RegExp(\"^https?:$\",\"i\"),i=0;i<o.length;i++)o[i].style.display=\"none\";for(i=0;i<a.length;i++)s=a[i],e.source===s.contentWindow&&(s.removeAttribute(\"style\"),\"height\"===t.message?(1e3<(r=parseInt(t.value,10))?r=1e3:~~r<200&&(r=200),s.height=r):\"link\"===t.message&&(r=new URL(s.getAttribute(\"src\")),n=new URL(t.value),c.test(n.protocol))&&n.host===r.host&&l.activeElement===s&&(d.top.location.href=t.value))}},d.addEventListener(\"message\",d.wp.receiveEmbedMessage,!1),l.addEventListener(\"DOMContentLoaded\",function(){for(var e,t,s=l.querySelectorAll(\"iframe.wp-embedded-content\"),r=0;r<s.length;r++)(t=(e=s[r]).getAttribute(\"data-secret\"))||(t=Math.random().toString(36).substring(2,12),e.src+=\"#?secret=\"+t,e.setAttribute(\"data-secret\",t)),e.contentWindow.postMessage({message:\"ready\",secret:t},\"*\")},!1)))}(window,document);\n\/\/# sourceURL=https:\/\/megagon.ai\/wp-includes\/js\/wp-embed.min.js\n<\/script>\n","thumbnail_url":"https:\/\/megagon.ai\/wp-content\/uploads\/2025\/02\/Publications.png","thumbnail_width":1600,"thumbnail_height":900,"description":"LCLM performance is more sensitive to how much information is packed into the context than to the length of that context. Moreover, tasks requiring aggregation of multiple facts across the input lead to noticeable performance drops, especially as complexity increases.These insights reveal a critical bottleneck in current LCLMs and point to where future work must focus: not just expanding context windows, but improving how models reason within them."}