\n\t<a id=\"✨-recverse-towards-faithful-simulation-of-human-shopping-behavior-🛒\" class=\"block pr-1.5 text-lg md:absolute md:p-1.5 md:opacity-0 md:group-hover:opacity-100 md:right-full\" href=\"#✨-recverse-towards-faithful-simulation-of-human-shopping-behavior-🛒\" rel=\"nofollow\">\n\t\t<span class=\"header-link\"><svg class=\"text-gray-500 hover:text-black dark:hover:text-gray-200 w-4\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" aria-hidden=\"true\" role=\"img\" width=\"1em\" height=\"1em\" preserveAspectRatio=\"xMidYMid meet\" viewBox=\"0 0 256 256\"><path d=\"M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z\" fill=\"currentColor\"></path></svg></span>\n\t</a>\n\t<span>\n\t\t✨ RecVerse: Towards Faithful Simulation of Human Shopping Behavior 🛒\n\t</span>\n</h1>\n<p>A GUI-grounded shopping agent that interacts with interfaces through screenshots 🖥️ like a real user, retains relevant information using a cognitively inspired hierarchical memory, and is trained end-to-end with trajectory-level reinforcement learning, producing simulated sessions that faithfully reflect how real people browse and shop online.</p>\n<p>📄 Paper: Towards Faithful Simulation of Human Shopping Behavior (arXiv:2608.20707)<br>🏛️ Affiliations: Renmin University of China · UCAS · NUS · Alibaba Group</p>\n<h2 class=\"relative group flex items-baseline\">\n\t<a id=\"🤔-why-should-you-care\" class=\"block pr-1.5 text-lg md:absolute md:p-1.5 md:opacity-0 md:group-hover:opacity-100 md:right-full\" href=\"#🤔-why-should-you-care\" rel=\"nofollow\">\n\t\t<span class=\"header-link\"><svg class=\"text-gray-500 hover:text-black dark:hover:text-gray-200 w-4\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" aria-hidden=\"true\" role=\"img\" width=\"1em\" height=\"1em\" preserveAspectRatio=\"xMidYMid meet\" viewBox=\"0 0 256 256\"><path d=\"M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z\" fill=\"currentColor\"></path></svg></span>\n\t</a>\n\t<span>\n\t\t🤔 Why should you care?\n\t</span>\n</h2>\n<p>Faithful user simulators power 🛍️ offline evaluation, counterfactual analysis, and RL recommender training — without burning online A/B traffic.<br>But current simulators still stumble on:<br>📚 Memory blow‑up across long sessions<br>🎯 Step‑wise imitation that reproduces noise and yields over‑active or overly passive behavior</p>\n<h2 class=\"relative group flex items-baseline\">\n\t<a id=\"💡recverses-two-big-ideas\" class=\"block pr-1.5 text-lg md:absolute md:p-1.5 md:opacity-0 md:group-hover:opacity-100 md:right-full\" href=\"#💡recverses-two-big-ideas\" rel=\"nofollow\">\n\t\t<span class=\"header-link\"><svg class=\"text-gray-500 hover:text-black dark:hover:text-gray-200 w-4\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" aria-hidden=\"true\" role=\"img\" width=\"1em\" height=\"1em\" preserveAspectRatio=\"xMidYMid meet\" viewBox=\"0 0 256 256\"><path d=\"M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z\" fill=\"currentColor\"></path></svg></span>\n\t</a>\n\t<span>\n\t\t💡RecVerse's Two Big Ideas\n\t</span>\n</h2>\n<h3 class=\"relative group flex items-baseline\">\n\t<a id=\"🧠-cognitiveinspired-hierarchical-memory\" class=\"block pr-1.5 text-lg md:absolute md:p-1.5 md:opacity-0 md:group-hover:opacity-100 md:right-full\" href=\"#🧠-cognitiveinspired-hierarchical-memory\" rel=\"nofollow\">\n\t\t<span class=\"header-link\"><svg class=\"text-gray-500 hover:text-black dark:hover:text-gray-200 w-4\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" aria-hidden=\"true\" role=\"img\" width=\"1em\" height=\"1em\" preserveAspectRatio=\"xMidYMid meet\" viewBox=\"0 0 256 256\"><path d=\"M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z\" fill=\"currentColor\"></path></svg></span>\n\t</a>\n\t<span>\n\t\t🧠 Cognitive‑Inspired Hierarchical Memory\n\t</span>\n</h3>\n<ul>\n<li>👁️ Working Memory\trecent screenshots + mindset (FIFO)</li>\n<li>📝 Episodic Memory\tsession‑level textual event trace</li>\n<li>❤️ Preference Memory\tdistilled long‑term user intent</li>\n</ul>\n<p><em>🪄 Memory‑as‑Action — the agent learns when and what to remember via RL.</em></p>\n<h2 class=\"relative group flex items-baseline\">\n\t<a id=\"🏹-trajectoryaligned-rl\" class=\"block pr-1.5 text-lg md:absolute md:p-1.5 md:opacity-0 md:group-hover:opacity-100 md:right-full\" href=\"#🏹-trajectoryaligned-rl\" rel=\"nofollow\">\n\t\t<span class=\"header-link\"><svg class=\"text-gray-500 hover:text-black dark:hover:text-gray-200 w-4\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" aria-hidden=\"true\" role=\"img\" width=\"1em\" height=\"1em\" preserveAspectRatio=\"xMidYMid meet\" viewBox=\"0 0 256 256\"><path d=\"M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z\" fill=\"currentColor\"></path></svg></span>\n\t</a>\n\t<span>\n\t\t🏹 Trajectory‑Aligned RL\n\t</span>\n</h2>\n<ul>\n<li>📊 Macro reward aligns action‑type distributions</li>\n<li>🔬 Micro reward uses a 3‑tier category tree for dense intent signals</li>\n<li>✅ Format reward keeps outputs valid</li>\n</ul>\n<h2 class=\"relative group flex items-baseline\">\n\t<a id=\"🗂️-usb-benchmark\" class=\"block pr-1.5 text-lg md:absolute md:p-1.5 md:opacity-0 md:group-hover:opacity-100 md:right-full\" href=\"#🗂️-usb-benchmark\" rel=\"nofollow\">\n\t\t<span class=\"header-link\"><svg class=\"text-gray-500 hover:text-black dark:hover:text-gray-200 w-4\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" aria-hidden=\"true\" role=\"img\" width=\"1em\" height=\"1em\" preserveAspectRatio=\"xMidYMid meet\" viewBox=\"0 0 256 256\"><path d=\"M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z\" fill=\"currentColor\"></path></svg></span>\n\t</a>\n\t<span>\n\t\t🗂️ USB Benchmark\n\t</span>\n</h2>\n<p><strong>The first interactive GUI benchmark supporting multi‑turn agentic RL.</strong></p>\n<ul>\n<li>🖼️ 5,274 trajectories </li>\n<li>🎬 69,842 actions </li>\n<li>🏷️ 90,095 items (41/517/2,256 categories) </li>\n<li>👤 5,222 users</li>\n</ul>\n<h2 class=\"relative group flex items-baseline\">\n\t<a id=\"🏆-results\" class=\"block pr-1.5 text-lg md:absolute md:p-1.5 md:opacity-0 md:group-hover:opacity-100 md:right-full\" href=\"#🏆-results\" rel=\"nofollow\">\n\t\t<span class=\"header-link\"><svg class=\"text-gray-500 hover:text-black dark:hover:text-gray-200 w-4\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\" aria-hidden=\"true\" role=\"img\" width=\"1em\" height=\"1em\" preserveAspectRatio=\"xMidYMid meet\" viewBox=\"0 0 256 256\"><path d=\"M167.594 88.393a8.001 8.001 0 0 1 0 11.314l-67.882 67.882a8 8 0 1 1-11.314-11.315l67.882-67.881a8.003 8.003 0 0 1 11.314 0zm-28.287 84.86l-28.284 28.284a40 40 0 0 1-56.567-56.567l28.284-28.284a8 8 0 0 0-11.315-11.315l-28.284 28.284a56 56 0 0 0 79.196 79.197l28.285-28.285a8 8 0 1 0-11.315-11.314zM212.852 43.14a56.002 56.002 0 0 0-79.196 0l-28.284 28.284a8 8 0 1 0 11.314 11.314l28.284-28.284a40 40 0 0 1 56.568 56.567l-28.285 28.285a8 8 0 0 0 11.315 11.314l28.284-28.284a56.065 56.065 0 0 0 0-79.196z\" fill=\"currentColor\"></path></svg></span>\n\t</a>\n\t<span>\n\t\t🏆 Results\n\t</span>\n</h2>\n<ul>\n<li>vs. STA (best GUI baseline): F1 +68%, HR +77%, HCO +41% 📈</li>\n<li>Head‑to‑head human eval: RecVerse beats STA in 92% of cases 🥇</li>\n<li>Falls within the high-fidelity behavioral band, while others remain in the over-active or over-passive zone.</li>\n</ul>\n","updatedAt":"2026-08-24T03:12:23.711Z","author":{"_id":"65acfb3a14e6582c30b4ce76","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65acfb3a14e6582c30b4ce76/RhEhePggBtyM0RIIqXQen.jpeg","fullname":"TangJiakai","name":"TangJiakai5704","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":3,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.6921426057815552},"editors":["TangJiakai5704"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/65acfb3a14e6582c30b4ce76/RhEhePggBtyM0RIIqXQen.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2608.20707","authors":[{"_id":"6a8bb6963d26296ea309193c","name":"Jiakai Tang","hidden":false},{"_id":"6a8bb6963d26296ea309193d","name":"Yan Mi","hidden":false},{"_id":"6a8bb6963d26296ea309193e","name":"Jing Yu","hidden":false},{"_id":"6a8bb6963d26296ea309193f","name":"Yang Zhang","hidden":false},{"_id":"6a8bb6963d26296ea3091940","name":"See-Kiong Ng","hidden":false},{"_id":"6a8bb6963d26296ea3091941","name":"Qi Cao","hidden":false},{"_id":"6a8bb6963d26296ea3091942","name":"Fei Sun","hidden":false},{"_id":"6a8bb6963d26296ea3091943","name":"Xu Chen","hidden":false},{"_id":"6a8bb6963d26296ea3091944","name":"Wen Chen","hidden":false},{"_id":"6a8bb6963d26296ea3091945","name":"Jian Wu","hidden":false},{"_id":"6a8bb6963d26296ea3091946","name":"Han Zhu","hidden":false},{"_id":"6a8bb6963d26296ea3091947","name":"Bo Zheng","hidden":false}],"publishedAt":"2026-08-21T00:00:00.000Z","submittedOnDailyAt":"2026-08-24T00:00:00.000Z","title":"Towards Faithful Simulation of Human Shopping Behavior","submittedOnDailyBy":{"_id":"65acfb3a14e6582c30b4ce76","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65acfb3a14e6582c30b4ce76/RhEhePggBtyM0RIIqXQen.jpeg","isPro":false,"fullname":"TangJiakai","user":"TangJiakai5704","type":"user","name":"TangJiakai5704"},"summary":"Simulating realistic user shopping behavior underpins offline evaluation and reinforcement learning in e-commerce scenarios. While recent LLM- and VLM-based simulators have made encouraging progress, reproducing a real browsing session remains difficult for two reasons. (i) Memory Challenge: a shopping session spans dozens of pages, yet existing agents either discard long-range observation histories, losing the evolving user state, or naively concatenate them, overwhelming the context window and even degrading simulation quality. (ii) Optimization Challenge: current user simulators are typically supervised to match each logged action via imitation or step-level rewards; the resulting sessions often display unrealistic patterns, such as over-exploration or excessive passivity, which per-step supervision can neither detect nor correct.\n To address the above challenges, we present RecVerse, a GUI-grounded simulation agent that perceives pages through screenshots and produces faithful multi-turn trajectories. For the memory challenge, RecVerse adopts a cognitive-inspired hierarchical memory: Working Memory for short-term focus, Episodic Memory for in-session traces, and Preference Memory for high-level intent, with memory updates treated as actions so that the agent adaptively learns when and what to memorize. For the optimization challenge, RecVerse is optimized with a trajectory-level RL objective that scores entire sessions, aligning both macro-level action-type distributions and micro-level shopping intent with real users. We further release USB (User Simulation Benchmark), an interactive e-commerce GUI trajectory dataset for multi-turn user simulation. Experiments show that RecVerse significantly outperforms existing baselines in both behavioral fidelity and intent consistency.","upvotes":3,"discussionId":"6a8bb6973d26296ea3091948","ai_summary":"RecVerse is a GUI-grounded agent that uses hierarchical memory and trajectory-level reinforcement learning to simulate realistic multi-turn e-commerce shopping sessions.","ai_keywords":["GUI-grounded simulation agent","hierarchical memory","working memory","episodic memory","preference memory","trajectory-level RL","user simulation benchmark"],"ai_summary_model":"thinkingmachines/Inkling-Small"},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"65acfb3a14e6582c30b4ce76","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/65acfb3a14e6582c30b4ce76/RhEhePggBtyM0RIIqXQen.jpeg","isPro":false,"fullname":"TangJiakai","user":"TangJiakai5704","type":"user"},{"_id":"63ac5701c21e60a3e9b58aa7","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/63ac5701c21e60a3e9b58aa7/g6EX7diOpuA94R2ab-rZC.png","isPro":true,"fullname":"Dipankar Sarkar","user":"dipankarsarkar","type":"user"},{"_id":"666a6486e400c5d6b90224b7","avatarUrl":"/avatars/6760ad58ae2e319537350497faaee4cf.svg","isPro":false,"fullname":"cici yue","user":"cicilya","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2608/2608.20707.md","query":{}}">

Get this paper in your agent:

Don't have the latest CLI?

Cite arxiv.org/abs/2608.20707 in a model README.md to link it from this page.

Cite arxiv.org/abs/2608.20707 in a dataset README.md to link it from this page.

Cite arxiv.org/abs/2608.20707 in a Space README.md to link it from this page.

Add this paper to a collection to link it from this page.