An online reasoning framework anticipates smartphone scams by analyzing streaming app-usage trajectories through self-evolving context management and on-policy self-distillation techniques.</p>\n","updatedAt":"2026-05-29T07:01:12.666Z","author":{"_id":"67e522d84ad46014c34efe90","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/PJZ9lrkQqOwjcKF7ecWHX.png","fullname":"gao","name":"snowleo135","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.8574650287628174},"editors":["snowleo135"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/PJZ9lrkQqOwjcKF7ecWHX.png"],"reactions":[],"isReport":false}},{"id":"6a1a417fb3470011372a83e2","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":359,"isUserFollowing":false},"createdAt":"2026-05-30T01:46:39.000Z","type":"comment","data":{"edited":false,"hidden":false,"latest":{"raw":"This is an automated message from the [Librarian Bot](https://huggingface.co/librarian-bots). I found the following papers similar to this paper. \n\nThe following papers were recommended by the Semantic Scholar API \n\n* [PreScam: A Benchmark for Predicting Scam Progression from Early Conversations](https://huggingface.co/papers/2605.12243) (2026)\n* [TRACES: Proactive Safety Auditing for Multi-Turn LLM Agents via Trajectory-State Modeling](https://huggingface.co/papers/2605.27690) (2026)\n* [StreamOV: Streaming Omni-Video Understanding via Evidence-Guided Memory and Response Triggering](https://huggingface.co/papers/2605.25621) (2026)\n* [Evaluating Memory Capability in Continuous Lifelog Scenario](https://huggingface.co/papers/2604.11182) (2026)\n* [A Low-Latency Fraud Detection Layer for Detecting Adversarial Interaction Patterns in LLM-Powered Agents](https://huggingface.co/papers/2605.01143) (2026)\n* [Follow the TRACE: Exploiting Post-Click Trajectories for Online Delayed Conversion Rate Prediction](https://huggingface.co/papers/2604.23197) (2026)\n* [HORIZON: A Benchmark for In-the-wild User Behaviour Modeling](https://huggingface.co/papers/2604.17259) (2026)\n\n\n Please give a thumbs up to this comment if you found it helpful!\n\n If you want recommendations for any Paper on Hugging Face checkout [this](https://huggingface.co/spaces/librarian-bots/recommend_similar_papers) Space\n\n You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: `@librarian-bot recommend`","html":"<p>This is an automated message from the <a href=\"https://huggingface.co/librarian-bots\">Librarian Bot</a>. I found the following papers similar to this paper. </p>\n<p>The following papers were recommended by the Semantic Scholar API </p>\n<ul>\n<li><a href=\"https://huggingface.co/papers/2605.12243\">PreScam: A Benchmark for Predicting Scam Progression from Early Conversations</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.27690\">TRACES: Proactive Safety Auditing for Multi-Turn LLM Agents via Trajectory-State Modeling</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.25621\">StreamOV: Streaming Omni-Video Understanding via Evidence-Guided Memory and Response Triggering</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.11182\">Evaluating Memory Capability in Continuous Lifelog Scenario</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2605.01143\">A Low-Latency Fraud Detection Layer for Detecting Adversarial Interaction Patterns in LLM-Powered Agents</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.23197\">Follow the TRACE: Exploiting Post-Click Trajectories for Online Delayed Conversion Rate Prediction</a> (2026)</li>\n<li><a href=\"https://huggingface.co/papers/2604.17259\">HORIZON: A Benchmark for In-the-wild User Behaviour Modeling</a> (2026)</li>\n</ul>\n<p> Please give a thumbs up to this comment if you found it helpful!</p>\n<p> If you want recommendations for any Paper on Hugging Face checkout <a href=\"https://huggingface.co/spaces/librarian-bots/recommend_similar_papers\">this</a> Space</p>\n<p> You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: <code><span class=\"SVELTE_PARTIAL_HYDRATER contents\" data-target=\"UserMention\" data-props=\"{"user":"librarian-bot"}\"><span class=\"inline-block\"><span class=\"contents\"><a href=\"/librarian-bot\">@<span class=\"underline\">librarian-bot</span></a></span> </span></span> recommend</code></p>\n","updatedAt":"2026-05-30T01:46:39.187Z","author":{"_id":"63d3e0e8ff1384ce6c5dd17d","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg","fullname":"Librarian Bot (Bot)","name":"librarian-bot","type":"user","isPro":false,"isHf":false,"isHfAdmin":false,"isMod":false,"followerCount":359,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.7377870082855225},"editors":["librarian-bot"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/1674830754237-63d3e0e8ff1384ce6c5dd17d.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2605.16363","authors":[{"_id":"6a181d546916a055bfaabdab","user":{"_id":"67e522d84ad46014c34efe90","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/PJZ9lrkQqOwjcKF7ecWHX.png","isPro":false,"fullname":"gao","user":"snowleo135","type":"user","name":"snowleo135"},"name":"Wenbo Gao","status":"claimed_verified","statusLastChangedAt":"2026-05-28T15:27:48.336Z","hidden":false},{"_id":"6a181d546916a055bfaabdac","name":"Songbai Tan","hidden":false},{"_id":"6a181d546916a055bfaabdad","name":"Zhongan Wang","hidden":false},{"_id":"6a181d546916a055bfaabdae","name":"Fei Shen","hidden":false},{"_id":"6a181d546916a055bfaabdaf","name":"Gang Xu","hidden":false},{"_id":"6a181d546916a055bfaabdb0","name":"Huiping Zhuang","hidden":false},{"_id":"6a181d546916a055bfaabdb1","name":"Yunyun Yang","hidden":false},{"_id":"6a181d546916a055bfaabdb2","name":"Ming Li","hidden":false},{"_id":"6a181d546916a055bfaabdb3","name":"Xiaofeng Zhu","hidden":false}],"publishedAt":"2026-05-09T00:00:00.000Z","submittedOnDailyAt":"2026-05-29T00:00:00.000Z","title":"ORACLE: Anticipating Scams from Partial Trajectories in Streaming App Usage","submittedOnDailyBy":{"_id":"67e522d84ad46014c34efe90","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/no-auth/PJZ9lrkQqOwjcKF7ecWHX.png","isPro":false,"fullname":"gao","user":"snowleo135","type":"user","name":"snowleo135"},"summary":"Smartphone scams are increasingly prevalent and typically manifest as multi-stage, cross-application processes with gradually emerging intent. Effective intervention thus requires anticipating scams before the intent becomes explicit. This is inherently challenging, as decisions must rely on partial trajectories with temporally distributed evidence. In this paper, we propose ORACLE Online Reasoning for Anticipating Cross-temporal Latent thrEats, the first agentic framework for early scam anticipation from streaming app-usage trajectories. To support this setting, we curate a real-world long-horizon benchmark of streaming app-usage trajectories, covering 12 scam types, spanning extended periods (15 days on average), involving diverse applications (95 apps), and interleaving normal and scam behaviors. To address fragmented evidence, we introduce a self-evolving context manager that adaptively consolidates entity-centric interactions over time, enabling more effective reconstruction of cross-temporal evidence from partial observations. To enhance sensitivity to latent early-stage signals, we propose an on-policy self-distillation scheme in which a teacher model, conditioned on summarized anti-scam reflections and clues by skills, supervises a student model without access to such reflections. This scheme thereby distills evidence-informed knowledge and improves recognition of emerging fraud patterns from partial trajectories. Experiments show that consistently improves early scam anticipation, yielding timely warnings while reducing false alerts in realistic streaming scenarios.","upvotes":0,"discussionId":"6a181d556916a055bfaabdb4","ai_summary":"An online reasoning framework anticipates smartphone scams by analyzing streaming app-usage trajectories through self-evolving context management and on-policy self-distillation techniques.","ai_keywords":["online reasoning","cross-temporal latent threats","streaming app-usage trajectories","self-evolving context manager","on-policy self-distillation","teacher-student model","fraud pattern recognition"]},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[],"acceptLanguages":["en"],"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2605/2605.16363.md"}">
ORACLE: Anticipating Scams from Partial Trajectories in Streaming App Usage
Published on May 9
· Submitted by gao on May 29 Abstract
An online reasoning framework anticipates smartphone scams by analyzing streaming app-usage trajectories through self-evolving context management and on-policy self-distillation techniques.
AI-generated summary
Smartphone scams are increasingly prevalent and typically manifest as multi-stage, cross-application processes with gradually emerging intent. Effective intervention thus requires anticipating scams before the intent becomes explicit. This is inherently challenging, as decisions must rely on partial trajectories with temporally distributed evidence. In this paper, we propose ORACLE Online Reasoning for Anticipating Cross-temporal Latent thrEats, the first agentic framework for early scam anticipation from streaming app-usage trajectories. To support this setting, we curate a real-world long-horizon benchmark of streaming app-usage trajectories, covering 12 scam types, spanning extended periods (15 days on average), involving diverse applications (95 apps), and interleaving normal and scam behaviors. To address fragmented evidence, we introduce a self-evolving context manager that adaptively consolidates entity-centric interactions over time, enabling more effective reconstruction of cross-temporal evidence from partial observations. To enhance sensitivity to latent early-stage signals, we propose an on-policy self-distillation scheme in which a teacher model, conditioned on summarized anti-scam reflections and clues by skills, supervises a student model without access to such reflections. This scheme thereby distills evidence-informed knowledge and improves recognition of emerging fraud patterns from partial trajectories. Experiments show that consistently improves early scam anticipation, yielding timely warnings while reducing false alerts in realistic streaming scenarios.
Community
An online reasoning framework anticipates smartphone scams by analyzing streaming app-usage trajectories through self-evolving context management and on-policy self-distillation techniques.
This is an automated message from the Librarian Bot. I found the following papers similar to this paper.
The following papers were recommended by the Semantic Scholar API
Please give a thumbs up to this comment if you found it helpful!
If you want recommendations for any Paper on Hugging Face checkout this Space
You can directly ask Librarian Bot for paper recommendations by tagging it in a comment: @librarian-bot recommend
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2605.16363 in a model README.md to link it from this page.
Cite arxiv.org/abs/2605.16363 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2605.16363 in a Space README.md to link it from this page.
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.