{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.09664", "kind": "paper", "url": "https://arxiv.org/abs/2609.09664"}], "authors": ["Hyojeong Yu", "Hyukhun Koh", "Minsung Kim", "Yunah Jang", "Kyomin Jung"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.09664", "source_url": "https://arxiv.org/abs/2609.09664"}, "categories": ["General AI", "Language & Knowledge", "Safety"], "description": {"en": "Evaluates personalized guidance through long-term conversation histories and evidence annotations, testing retrieval accuracy and memory-grounded reasoning."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.09664", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.09664", "title": "PRAGMA"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2609.09664", "modality": null, "name": "PRAGMA", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "true", "origin_record_id": "bm_pragma_88c92ac1", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "5c340b0ce20ca00a39b0af73c8039300743b170d7553360f047e8b8cb8cd249d", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-reviewed", "reviewed_at": "2026-09-11T00:52:53.785364Z", "source_url": "https://arxiv.org/abs/2609.09664"}, "publisher": null, "released": "2026-09-09", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.09664", "source_url": "https://arxiv.org/abs/2609.09664"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-09664", "source": "claire_radar", "source_benchmark_id": "2609.09664", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-09-10", "attemptedAt": "2026-09-10T12:24:43.080049Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.09664", "observedAt": "2026-09-09T12:31:07.558041Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-09-10T12:24:43.080049Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "Personalized assistants and memory systems are gaining attention, but the lack of immediate artifacts may slow initial adoption.", "score": 38}, "capabilities": [], "confidence": "High", "construction": "Transform Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-11T00:52:53.785364Z", "inputHash": "19e1090eb8a70475023ddc1b124d992f518280bc6bb7e1036ff44fcf6b86d7a9", "model": "deepseek-v4-pro"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "we introduce PRAGMA, a benchmark for evaluating personalized guidance in long-term conversations", "canonicalNameSource": "abstract", "decisionReason": "Benchmark task and experimental protocol are described, though no direct data/code link; clear evaluation object.", "evaluationEvidence": {"reuse": "Experiments across retrieval systems, memory systems, and long-context models", "scoring": "Experiments across retrieval systems, memory systems, and long-context models reveal that current systems struggle both to recover the appropriate conversational evidence and to effectively use it for personalized guidance.", "task": "evaluating personalized guidance in long-term conversations; contains curated longitudinal conversation histories, evidence annotations, and guidance scenarios grounded in evolving user contexts and incorrect user assumptions."}, "model": "deepseek-v4-pro", "reviewedAt": "2026-09-11T00:52:53.785364Z", "state": "ai-reviewed"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates personalized guidance through long-term conversation histories and evidence annotations, testing retrieval accuracy and memory-grounded reasoning.", "displayEligible": true, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["exact coined title identity tied to benchmark evidence", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "To study this challenge, we introduce PRAGMA, a benchmark for evaluating personalized guidance in long-term conversations."}, "familyId": "bmf_6f8bdbe23bec", "firstSeenAt": "2026-09-10", "heat": null, "id": "bm_pragma_88c92ac1", "indexedAt": "2026-09-10T00:47:26.889471Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2609.09664", "pdf": "https://arxiv.org/pdf/2609.09664", "project": null, "report": "https://arxiv.org/abs/2609.09664"}, "metrics": [], "motivation": "Large language models (LLMs) are increasingly deployed as personalized assistants that interact with users over extended periods of time.", "name": "PRAGMA", "oneLine": "Evaluates personalized guidance through long-term conversation histories and evidence annotations, testing retrieval accuracy and memory-grounded reasoning.", "paperTitle": "PRAGMA: Evaluating Personalized Guidance with Memory Alignment in Lifelong Conversations", "primaryDomain": "General AI", "publication": {"evidence": "Accepted to EMNLP 2026", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2609.09664", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "EMNLP 2026", "verifiedAt": "2026-09-10T00:47:26.889471Z"}, "publications": [], "ranking": {"30d": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 30}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}, "90d": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 90}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}, "today": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": 1, "historyDate": "2026-09-09", "rank": null, "score": null, "windowDays": 1}, "level": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "llmAttentionForecast": {"maxContribution": 0.25, "percentile": 0.57, "role": "bonus", "value": 57}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": 14}, "method": "attention-ranking-v13", "rank": null, "score": 14}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-09-09", "source": {"authors": ["Hyojeong Yu", "Hyukhun Koh", "Minsung Kim", "Yunah Jang", "Kyomin Jung"], "categories": ["cs.AI"], "id": "2609.09664", "title": "PRAGMA: Evaluating Personalized Guidance with Memory Alignment in Lifelong Conversations", "type": "arxiv", "url": "https://arxiv.org/abs/2609.09664"}, "sourceUpdatedAt": "2026-09-09T03:31:47Z", "topics": ["Safety"], "venueAttempts": [{"decisionRaw": "Accepted to EMNLP 2026", "evidence": [{"level": "author-claim", "observedAt": "2026-09-10T00:47:26.889471Z", "rawValue": "Accepted to EMNLP 2026", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2609.09664"}], "reviewStatus": "accepted", "venueName": "EMNLP 2026"}], "whyItMatters": "Shifts memory evaluation from factual recall to practical guidance, uncovering deficits in current memory systems for user-specific reasoning."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
