{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:ZKTZZOABMUIF7Q6PV2P7IVVB3W","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"957971901f1162d5a2926870ced3513360cb383fe941ac3900b98181425aed96","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T14:45:59Z","title_canon_sha256":"46d10ca4b7de8cd460b1163630d4023d513748f5207f7f6c6efdee4d2be9c368"},"schema_version":"1.0","source":{"id":"2605.28556","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.28556","created_at":"2026-05-28T02:04:56Z"},{"alias_kind":"arxiv_version","alias_value":"2605.28556v1","created_at":"2026-05-28T02:04:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.28556","created_at":"2026-05-28T02:04:56Z"},{"alias_kind":"pith_short_12","alias_value":"ZKTZZOABMUIF","created_at":"2026-05-28T02:04:56Z"},{"alias_kind":"pith_short_16","alias_value":"ZKTZZOABMUIF7Q6P","created_at":"2026-05-28T02:04:56Z"},{"alias_kind":"pith_short_8","alias_value":"ZKTZZOAB","created_at":"2026-05-28T02:04:56Z"}],"graph_snapshots":[{"event_id":"sha256:82b62ea8246fa3dcad18ab615a20f9ec492f857b42a8e2847458b991b2f9bf70","target":"graph","created_at":"2026-05-28T02:04:56Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2605.28556/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As agent capabilities advance, existing benchmarks, such as $\\tau^2$-Bench, are becoming increasingly saturated. Yet constructing new benchmark tasks remains complex, costly, and labor-intensive. Moreover, the standard approach, in which scenarios are first written in natural language and then mapped to tool sequences, captures only a narrow subset of the tool-use patterns agents exercise. In this paper, we address these problems by reversing the task construction process. We propose TASTE: Task Synthesis from Tool Sequence Evolution, an automatic method that generates challenging tasks with b","authors_text":"Asaf Yehudai, Michal Shmueli-Scheuer, Nitay Calderon, Roi Reichert, Tomer Keren, Yotam Perlitz","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T14:45:59Z","title":"A Matter of TASTE: Improving Coverage and Difficulty of Agent Benchmarks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.28556","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:0618ed4413c52b9600bcc874dc30df35d616f3e2417f99ef577beda1d4c3042f","target":"record","created_at":"2026-05-28T02:04:56Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"957971901f1162d5a2926870ced3513360cb383fe941ac3900b98181425aed96","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T14:45:59Z","title_canon_sha256":"46d10ca4b7de8cd460b1163630d4023d513748f5207f7f6c6efdee4d2be9c368"},"schema_version":"1.0","source":{"id":"2605.28556","kind":"arxiv","version":1}},"canonical_sha256":"caa79cb80165105fc3cfae9ff456a1dd98a78ba1555f70ff570e1d4221e9deb3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"caa79cb80165105fc3cfae9ff456a1dd98a78ba1555f70ff570e1d4221e9deb3","first_computed_at":"2026-05-28T02:04:56.287593Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-28T02:04:56.287593Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ANz/CX8fZDlajC33hexqSwt7LdsIOa+95t8zdcQ0tOBka7IDF2+ZduVsbXd0pG9JIjz+qKn4CWVeSemeE60QAA==","signature_status":"signed_v1","signed_at":"2026-05-28T02:04:56.288004Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.28556","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0618ed4413c52b9600bcc874dc30df35d616f3e2417f99ef577beda1d4c3042f","sha256:82b62ea8246fa3dcad18ab615a20f9ec492f857b42a8e2847458b991b2f9bf70"],"state_sha256":"4929092bc960a44e7d3808fc63b844871da0aaa22fb29aac6b2f2702039e8da5"}