{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2607.20498", "kind": "paper", "url": "https://arxiv.org/abs/2607.20498"}, {"id": "gh:zai-org/aise-bench", "kind": "repo", "url": "https://github.com/zai-org/AISE-Bench"}, {"kind": "website", "url": "https://aise-bench.github.io/"}], "authors": ["Fanjin Zhang", "Zhengyang Wang", "Ruixuan Huang", "Kefan Zhang", "Amy Xin", "Yuanchun Wang", "Shu Zhao", "Evgeny Kharlamov", "Jie Tang", "Juanzi Li"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2607.20498", "source_url": "https://arxiv.org/abs/2607.20498"}, "categories": ["Factuality", "General AI", "Language & Knowledge", "cs.AI"], "description": {"en": "AISE-Bench is a benchmark for evaluating multi-step API-using LLM agents in information seeking on academic knowledge graphs. It comprises 1,133 QA pairs with API trajectories, validated parameters, and grounded answers, and evaluates answer quality, reference grounding, API-planning correctness, and execution success."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2607.20498", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2607.20498", "title": "AISE-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2607.20498", "modality": null, "name": "AISE-Bench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "unknown", "origin_record_id": "bm_aise-bench_0ef6ecf5", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "2173b39d714b64c983bece7a07ed771e79979d6b61a89c27b9f228cb1d2d5c85", "relation": "introduces", "review_state": "unreviewed", "source_url": "https://arxiv.org/abs/2607.20498"}, "publisher": null, "released": "2026-06-16", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2607.20498", "source_url": "https://arxiv.org/abs/2607.20498"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2607-20498", "source": "claire_radar", "source_benchmark_id": "2607.20498", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Machine Generated", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": "https://github.com/zai-org/AISE-Bench", "githubScope": "benchmark_repo", "githubStars": 6, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": 0, "hfPaperUrl": "https://huggingface.co/papers/2607.20498", "observedAt": "2026-10-04T15:31:48.030916Z", "signalStatus": {"githubStars": {"observedAt": "2026-10-04T15:31:48.030916Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "lastSuccessfulDate": "2026-10-01", "state": "stale"}}, "status": "fresh"}, "capabilities": ["Factuality"], "confidence": "High", "construction": "Original Synthetic", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-08-20T15:22:10.044731Z", "inputHash": "f2c2746249a7d4f2ece9a449573b9de47792cbb8c3d25e47f8b271e7cf14094e", "model": "deepseek-v4-flash"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "AISE-Bench is a benchmark for evaluating multi-step API-using LLM agents in information seeking on academic knowledge graphs. It comprises 1,133 QA pairs with API trajectories, validated parameters, and grounded answers, and evaluates answer quality, reference grounding, API-planning correctness, and execution success.", "domainCuration": {"method": "rules-v1", "state": "auto"}, "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "evaluation protocol evidence", "public artifact URL"], "snippet": "We introduce AISE-Bench, a real-world, full-cycle annotated benchmark for information seeking on academic knowledge graphs."}, "familyId": "bmf_9462adb5901c", "firstSeenAt": "2026-08-19", "githubIndex": {"candidateCount": 2, "indexedAt": "2026-09-17T04:40:34.111665+00:00", "repositories": [{"evidence": [{"basis": "linked-from-project", "linkedUrl": "https://github.com/zai-org/AISE-Bench", "url": "https://aise-bench.github.io/"}], "scope": "benchmark_repo", "url": "https://github.com/zai-org/AISE-Bench", "verification": "source-linked"}], "status": "linked"}, "heat": null, "id": "bm_aise-bench_0ef6ecf5", "indexedAt": "2026-08-19T11:05:13.395391Z", "industrySectors": [], "links": {"code": "https://github.com/zai-org/AISE-Bench", "data": null, "hfPaper": "https://huggingface.co/papers/2607.20498", "paper": "https://arxiv.org/abs/2607.20498", "pdf": "https://arxiv.org/pdf/2607.20498", "project": "https://aise-bench.github.io/", "report": "https://arxiv.org/abs/2607.20498"}, "metrics": [], "motivation": "Large language models (LLMs) augmented with tools are emerging as autonomous agents capable of using Web engine, APIs, and code to solve complex, long-horizon tasks.", "name": "AISE-Bench", "oneLine": "AISE-Bench is a benchmark for evaluating multi-step API-using LLM agents in information seeking on academic knowledge graphs. It comprises 1,133 QA pairs with API trajectories, validated parameters, and grounded answers, and evaluates answer quality, reference grounding, API-planning correctness, and execution success.", "paperTitle": "AISE-Bench: A Full-Cycle Curated Benchmark for Information Seeking on Academic Knowledge Graphs", "primaryDomain": "General AI", "publication": {"evidence": "Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD '26), August 09-13, 2026, Jeju Island, Republic of Korea", "evidenceLevel": "strong-author-metadata", "evidenceUrl": "https://arxiv.org/abs/2607.20498", "source": "arxiv-journal-reference", "status": "publication_reported", "venue": "Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD '26), August 09-13, 2026, Jeju Island, Republic of Korea", "verifiedAt": "2026-10-14T01:46:58.058199Z"}, "publications": [{"evidence": [{"level": "strong-author-metadata", "observedAt": "2026-10-14T01:46:58.058199Z", "rawValue": "Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD '26), August 09-13, 2026, Jeju Island, Republic of Korea", "sourceType": "arxiv-journal-reference", "sourceUrl": "https://arxiv.org/abs/2607.20498"}], "publicationStatus": "published", "venueName": "Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD '26), August 09-13, 2026, Jeju Island, Republic of Korea"}], "publishers": [{"name": "AISE-Bench Team", "organizationType": "academic-lab", "role": "benchmark-publisher", "sourceUrl": "https://aise-bench.github.io/"}], "ranking": {}, "readiness": "Runnable", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-06-16", "source": {"authors": ["Fanjin Zhang", "Zhengyang Wang", "Ruixuan Huang", "Kefan Zhang", "Amy Xin", "Yuanchun Wang", "Shu Zhao", "Evgeny Kharlamov", "Jie Tang", "Juanzi Li"], "categories": ["cs.AI"], "id": "2607.20498", "title": "AISE-Bench: A Full-Cycle Curated Benchmark for Information Seeking on Academic Knowledge Graphs", "type": "arxiv", "url": "https://arxiv.org/abs/2607.20498"}, "sourceUpdatedAt": "2026-06-16T07:59:46Z", "topics": ["cs.AI"], "venueAttempts": [], "whyItMatters": "Existing benchmarks for tool-using agents on academic graphs rely on synthetic or narrow tasks. AISE-Bench addresses this gap with real-world, full-cycle annotated data, enabling quantitative assessment of stepwise correctness, grounded summarization, and traceable reasoning in complex API workflows. It provides a challenging testbed for improving agent reliability in realistic information-seeking scenarios."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
