{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:E2EXS5LATKSRIS2AAWOTFIPYP5","short_pith_number":"pith:E2EXS5LA","canonical_record":{"source":{"id":"2605.29656","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T09:19:50Z","cross_cats_sorted":[],"title_canon_sha256":"c025c3327f7f02bb2fd8ad2c2ec10931632d27b46932a22d3f5481a3ec11d41e","abstract_canon_sha256":"d92c50446ffa57c848a18eff005be6c75157102d3af69c3a4f5f51a884ce4795"},"schema_version":"1.0"},"canonical_sha256":"26897975609aa5144b40059d32a1f87f458b738733ab7f434e1f036539d4c643","source":{"kind":"arxiv","id":"2605.29656","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.29656","created_at":"2026-05-29T01:05:53Z"},{"alias_kind":"arxiv_version","alias_value":"2605.29656v1","created_at":"2026-05-29T01:05:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.29656","created_at":"2026-05-29T01:05:53Z"},{"alias_kind":"pith_short_12","alias_value":"E2EXS5LATKSR","created_at":"2026-05-29T01:05:53Z"},{"alias_kind":"pith_short_16","alias_value":"E2EXS5LATKSRIS2A","created_at":"2026-05-29T01:05:53Z"},{"alias_kind":"pith_short_8","alias_value":"E2EXS5LA","created_at":"2026-05-29T01:05:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:E2EXS5LATKSRIS2AAWOTFIPYP5","target":"record","payload":{"canonical_record":{"source":{"id":"2605.29656","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T09:19:50Z","cross_cats_sorted":[],"title_canon_sha256":"c025c3327f7f02bb2fd8ad2c2ec10931632d27b46932a22d3f5481a3ec11d41e","abstract_canon_sha256":"d92c50446ffa57c848a18eff005be6c75157102d3af69c3a4f5f51a884ce4795"},"schema_version":"1.0"},"canonical_sha256":"26897975609aa5144b40059d32a1f87f458b738733ab7f434e1f036539d4c643","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T01:05:53.884436Z","signature_b64":"YprWlLvzbNz/JZ34MBPCDmSiIBysIgShhUNzA8SKwsdbMHywMXD5AUw0BPb0ghGzW+oBjFRy4cSTNSeq4RJXCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"26897975609aa5144b40059d32a1f87f458b738733ab7f434e1f036539d4c643","last_reissued_at":"2026-05-29T01:05:53.883966Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T01:05:53.883966Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.29656","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-29T01:05:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eAa0Rm3Tga9ccouLit5c78j7TXNTJ0TMBHs+sLlf55kbhHZmhQGfDAw9DQ7drfvqmP2gOkkCtVN0xJkyjG88CA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T13:26:54.009285Z"},"content_sha256":"1163be399113e628ee5aadbca8ffeea43bb798261f5b5975eaf92914c681bcc4","schema_version":"1.0","event_id":"sha256:1163be399113e628ee5aadbca8ffeea43bb798261f5b5975eaf92914c681bcc4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:E2EXS5LATKSRIS2AAWOTFIPYP5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TRACE: Toulmin-based Reasoning Assessment through Constructive Elements for LLM CoT Evaluation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Heyoung Yang, Yundong Kim","submitted_at":"2026-05-28T09:19:50Z","abstract_excerpt":"Evaluating open-ended outputs from large language models (LLMs) remains challenging due to the absence of ground truth. Existing metrics rely on final-answer accuracy or surface-level statistics, leaving the reasoning process itself unexamined. We introduce TRACE (Toulmin-based Reasoning Assessment through Constructive Elements), a metric that analyzes Chain-of-Thought (CoT) reasoning processes. Rather than judging outcomes, TRACE inspects how arguments are constructed by integrating Toulmin's argumentation theory with Flavell's metacognitive framework to assess reasoning structure. Experiment"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.29656","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.29656/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-29T01:05:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tseOSUFjldOytTA0bPXSecKFyKwF5Xk8dO+qIKzb3p1NusmYC+AKq2jQjuLOk2pQUZZV5rdkYvhsSNtZRVAOAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T13:26:54.009673Z"},"content_sha256":"6e3cbc1ae7c0f0ec991dd24e4133c6fa2a14b414dcc3b7fd4d07a7fe803bb665","schema_version":"1.0","event_id":"sha256:6e3cbc1ae7c0f0ec991dd24e4133c6fa2a14b414dcc3b7fd4d07a7fe803bb665"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/E2EXS5LATKSRIS2AAWOTFIPYP5/bundle.json","state_url":"https://pith.science/pith/E2EXS5LATKSRIS2AAWOTFIPYP5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/E2EXS5LATKSRIS2AAWOTFIPYP5/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-29T13:26:54Z","links":{"resolver":"https://pith.science/pith/E2EXS5LATKSRIS2AAWOTFIPYP5","bundle":"https://pith.science/pith/E2EXS5LATKSRIS2AAWOTFIPYP5/bundle.json","state":"https://pith.science/pith/E2EXS5LATKSRIS2AAWOTFIPYP5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/E2EXS5LATKSRIS2AAWOTFIPYP5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:E2EXS5LATKSRIS2AAWOTFIPYP5","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":"d92c50446ffa57c848a18eff005be6c75157102d3af69c3a4f5f51a884ce4795","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T09:19:50Z","title_canon_sha256":"c025c3327f7f02bb2fd8ad2c2ec10931632d27b46932a22d3f5481a3ec11d41e"},"schema_version":"1.0","source":{"id":"2605.29656","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.29656","created_at":"2026-05-29T01:05:53Z"},{"alias_kind":"arxiv_version","alias_value":"2605.29656v1","created_at":"2026-05-29T01:05:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.29656","created_at":"2026-05-29T01:05:53Z"},{"alias_kind":"pith_short_12","alias_value":"E2EXS5LATKSR","created_at":"2026-05-29T01:05:53Z"},{"alias_kind":"pith_short_16","alias_value":"E2EXS5LATKSRIS2A","created_at":"2026-05-29T01:05:53Z"},{"alias_kind":"pith_short_8","alias_value":"E2EXS5LA","created_at":"2026-05-29T01:05:53Z"}],"graph_snapshots":[{"event_id":"sha256:6e3cbc1ae7c0f0ec991dd24e4133c6fa2a14b414dcc3b7fd4d07a7fe803bb665","target":"graph","created_at":"2026-05-29T01:05:53Z","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/2605.29656/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Evaluating open-ended outputs from large language models (LLMs) remains challenging due to the absence of ground truth. Existing metrics rely on final-answer accuracy or surface-level statistics, leaving the reasoning process itself unexamined. We introduce TRACE (Toulmin-based Reasoning Assessment through Constructive Elements), a metric that analyzes Chain-of-Thought (CoT) reasoning processes. Rather than judging outcomes, TRACE inspects how arguments are constructed by integrating Toulmin's argumentation theory with Flavell's metacognitive framework to assess reasoning structure. Experiment","authors_text":"Heyoung Yang, Yundong Kim","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T09:19:50Z","title":"TRACE: Toulmin-based Reasoning Assessment through Constructive Elements for LLM CoT Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.29656","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":"sha256:1163be399113e628ee5aadbca8ffeea43bb798261f5b5975eaf92914c681bcc4","target":"record","created_at":"2026-05-29T01:05:53Z","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":"d92c50446ffa57c848a18eff005be6c75157102d3af69c3a4f5f51a884ce4795","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-28T09:19:50Z","title_canon_sha256":"c025c3327f7f02bb2fd8ad2c2ec10931632d27b46932a22d3f5481a3ec11d41e"},"schema_version":"1.0","source":{"id":"2605.29656","kind":"arxiv","version":1}},"canonical_sha256":"26897975609aa5144b40059d32a1f87f458b738733ab7f434e1f036539d4c643","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"26897975609aa5144b40059d32a1f87f458b738733ab7f434e1f036539d4c643","first_computed_at":"2026-05-29T01:05:53.883966Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-29T01:05:53.883966Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"YprWlLvzbNz/JZ34MBPCDmSiIBysIgShhUNzA8SKwsdbMHywMXD5AUw0BPb0ghGzW+oBjFRy4cSTNSeq4RJXCg==","signature_status":"signed_v1","signed_at":"2026-05-29T01:05:53.884436Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.29656","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1163be399113e628ee5aadbca8ffeea43bb798261f5b5975eaf92914c681bcc4","sha256:6e3cbc1ae7c0f0ec991dd24e4133c6fa2a14b414dcc3b7fd4d07a7fe803bb665"],"state_sha256":"738c5ac0e214bdd9aead5436fe4eaf29c7ebdc90b660349ba7ddd56741bec8d7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"k09dZSivWCudVKYXiR+9hkJOMJJxNKe/2Dq90Gk/8SHSKZFM0hWJImbYEWGrIwlpaL4vg9IwoxIcKLAE2iHOAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-29T13:26:54.011731Z","bundle_sha256":"9279ac28bddc2c3ef25cb514c0939e4d75d1c442a39e30660db4be4f98ba4197"}}