[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-85720-en":3,"doc-seo-85720-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},85720,4398048950312,"Violet","https://ap-avatar.wpscdn.com/avatar/400002538284de19e3c?_k=1778320343897328908",8,"Research & Report","Agentic Context Learning with Self-Discovered Specification","Context learning tasks require LLMs to learn and apply novel, task-specific knowledge from complex contexts that are absent from pretraining. Empirical evidence shows that, even on CL-Bench, frontier models achieve under 24% strict success and many baselines over direct full-context prompting gain little. Failure analysis attributes performance loss to local specification acquisition: domain formats, behavioral rules, and completeness conditions. A new intervention, PSCI, extracts and enforces private specification contracts via checking and repair, reaching state-of-the-art 28.14% success and consistent improvements via ablations and replication.","arXiv :2607 .09794v 1 [ cs .AI] 9 Jul 2026  \nAGENTIC CONTEXT LEARNING WITH SELFDISCOVERED SPECIFICATION  \nJike Zhong 1∗†, Ming Li2∗, Yuxiang Lai3∗, Ziyan Yang 1 , Jingyu Xie 1 , Jihyung Kil4 , Zheda Mai5 , Shao-Yuan Lo6 , Xiang Ren 1 , Konstantinos Psounis 1 , Yuanyuan Lei2  \n1University of Southern California 2University of Florida 3Emory University  \n4Adobe Research 5The Ohio State University 6National Taiwan University  \nABSTRACT  \nContext learning is an emerging inference-time task where LLMs must learn and apply novel, task-specific knowledge from intricate contexts absent from pretraining; even frontier models score under 24% task success. In this work, we conduct a comprehensive empirical study to understand why this setting remains difficult. A natural hypothesis is that failures stem from content access; yet across twelve retrieval, reflection, and verification baselines on CL-Bench, an extensive context learning benchmark, we find limited gains over direct full-context prompting. Further failure analysis reveals a key finding: unlike typical longcontext tasks such as long document understanding, context learning requires not only recovering local content but also acquiring local specifications that are often unspecified in the query but distributed across the context: domain-specific formats, local rules, and completeness conditions. Across all 31,592 rubric items, we find that 55.4% clearly evaluate specification acquisition, while only 22.6% evaluate content acquisition. Moreover, despite 76.7% of specifications being unspecified in the user query, 95.5% are traceable to the context, indicating these are learnable obligations rather than hidden requirements. To validate this diagnosis, we design a deliberately simple intervention PSCI (private specification-contract induction) which extracts local specifications and enforces them through adversarial checking and repair; PSCI achieves state-of-the-art 28.14% with GPT-5.1 (+5 .59 pp absolute and +24 .8% relative) on CL-Bench, replicated on Qwen3 .5-27B (+5 .28 pp) and Gemini 3 Pro (+6 .17 pp) . Seventeen ablations further isolate the role of taskspecific specifications. Overall, our results suggest context learning hinges on not only content acquisition but also specification acquisition.  \n1 INTRODUCTION  \nLarge language models are increasingly deployed in settings where the prompt is not merely evidence to retrieve from, but defines a novel operating environment the model must learn before it can act: a product manual, workflow policy, regulatory code, API specification, or experimental dataset, often spanning tens of thousands of tokens and specifying the rules of a local domain. Consider the task of detecting anomalies in an incoming event stream: the user prompt contains a 40K-token context followed by a simple query asking the model to record and log anomalies. The context defines not only the event semantics needed to detect anomalies, but also local specifications that determine what counts as a valid log: for instance, if a reserved event ID is skipped, the log must emit the exact flag Sequence_Gap_Warning. In this case, identifying the missing ID is the query-relevant content; emitting the prescribed flag is a validity-relevant specification. A response that detects the anomaly but omits the flag is still wrong. Thus, the challenge is not merely retrieving and reasoning over along prompt, which is the focus of traditional long-context benchmarks (Liu et al., 2024; Hsieh et al., 2024; Bai et al., 2024), but acquiring the local specifications by which the answer is validated.  \nDespite strong performance across many language tasks, frontier LLMs remain fragile on context learning: the best model in the original CL-Bench evaluation scores below 24% strict task success.  \n*  \nEqual contribution.  \n†Correspondence to: [jikezhon@usc.edu](jikezhon@usc.edu).  \nFigure 1: Left: Context learning decomposes into content acquisition (what the answer says) and specificat","cbCaijwEQk5STWUs","https://ap.wps.com/l/cbCaijwEQk5STWUs","pdf",1728738,1,27,"English","en",105,"# Abstract\n# Introduction\n## Problem setting and motivation\n## CL-Bench benchmark\n## Empirical study and baseline evaluation\n# Failure analysis\n## Content vs. specification acquisition\n## Rubric-level patterns\n# Proposed method\n## PSCI: specification-contract induction\n# Experiments\n## Results and replications\n# Ablations and conclusions","[{\"question\":\"What makes context learning different from traditional long-context benchmarks?\",\"answer\":\"Context learning requires acquiring locally specified behavioral rules and validity conditions in addition to extracting relevant content. Even correct detection without required flags or formatting can fail the rubric.\"},{\"question\":\"Why do many retrieval or restructuring baselines perform poorly on CL-Bench?\",\"answer\":\"Across twelve methods, gains over direct full-context prompting are limited, indicating that content access alone does not resolve the core difficulty. Rubric-level failures show missing compliance requirements rather than missing facts.\"},{\"question\":\"How does PSCI improve performance in context learning?\",\"answer\":\"PSCI extracts local specifications from the task and context, then enforces them using adversarial checking and repair. This creates a private specification contract that guides generation to meet unseen rubric constraints.\"}]",1784205791,68,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"agentic-context-learning-with-self-discovered-specification","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/agentic-context-learning-with-self-discovered-specification/85720/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What makes context learning different from traditional long-context benchmarks?","Question",{"text":75,"@type":76},"Context learning requires acquiring locally specified behavioral rules and validity conditions in addition to extracting relevant content. Even correct detection without required flags or formatting can fail the rubric.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why do many retrieval or restructuring baselines perform poorly on CL-Bench?",{"text":80,"@type":76},"Across twelve methods, gains over direct full-context prompting are limited, indicating that content access alone does not resolve the core difficulty. Rubric-level failures show missing compliance requirements rather than missing facts.",{"name":82,"@type":73,"acceptedAnswer":83},"How does PSCI improve performance in context learning?",{"text":84,"@type":76},"PSCI extracts local specifications from the task and context, then enforces them using adversarial checking and repair. This creates a private specification contract that guides generation to meet unseen rubric constraints.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]