{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:X7YDW5XK3DOYEAWIXIKAZEKWXR","short_pith_number":"pith:X7YDW5XK","canonical_record":{"source":{"id":"2412.05288","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.SE","submitted_at":"2024-11-21T11:20:48Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"df53d77587772e4a4b8b1ba5ec018653539b98aacb7b8493e63b08ead4b64f06","abstract_canon_sha256":"09a51479a820c58c959baa7b32b2f0b53c8e8b95311ff6ef43289f609af79908"},"schema_version":"1.0"},"canonical_sha256":"bff03b76ead8dd8202c8ba140c9156bc6568f38793f5e4b99ff82e47b1e2f4aa","source":{"kind":"arxiv","id":"2412.05288","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.05288","created_at":"2026-07-05T09:46:00Z"},{"alias_kind":"arxiv_version","alias_value":"2412.05288v1","created_at":"2026-07-05T09:46:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.05288","created_at":"2026-07-05T09:46:00Z"},{"alias_kind":"pith_short_12","alias_value":"X7YDW5XK3DOY","created_at":"2026-07-05T09:46:00Z"},{"alias_kind":"pith_short_16","alias_value":"X7YDW5XK3DOYEAWI","created_at":"2026-07-05T09:46:00Z"},{"alias_kind":"pith_short_8","alias_value":"X7YDW5XK","created_at":"2026-07-05T09:46:00Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:X7YDW5XK3DOYEAWIXIKAZEKWXR","target":"record","payload":{"canonical_record":{"source":{"id":"2412.05288","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.SE","submitted_at":"2024-11-21T11:20:48Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"df53d77587772e4a4b8b1ba5ec018653539b98aacb7b8493e63b08ead4b64f06","abstract_canon_sha256":"09a51479a820c58c959baa7b32b2f0b53c8e8b95311ff6ef43289f609af79908"},"schema_version":"1.0"},"canonical_sha256":"bff03b76ead8dd8202c8ba140c9156bc6568f38793f5e4b99ff82e47b1e2f4aa","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:46:00.781498Z","signature_b64":"jcLbZveZZZdgMZqZ9MzTUyARl8uuk3tw4HTQJW298FCNONh1+qIBb0U1QsBRCx7yrx/7HTgSq0/0d15w6wxiAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bff03b76ead8dd8202c8ba140c9156bc6568f38793f5e4b99ff82e47b1e2f4aa","last_reissued_at":"2026-07-05T09:46:00.781147Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:46:00.781147Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.05288","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":"2026-07-05T09:46:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4jUKbMZSmgjpskduDAeaKvJy1sjdVmhotpm7y4pcOZeuTsDokr1TBSaUdRMc4ulSIPdtt9Y+Okc+J2+WbG9PCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T03:04:42.348961Z"},"content_sha256":"704b310da832358772b04d38e70ae59793f4c99223fa621b603dafd9a77a2bfa","schema_version":"1.0","event_id":"sha256:704b310da832358772b04d38e70ae59793f4c99223fa621b603dafd9a77a2bfa"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:X7YDW5XK3DOYEAWIXIKAZEKWXR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"StackEval: Benchmarking LLMs in Coding Assistance","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.SE","authors_text":"Dogu Araci, Nidhish Shah, Zulkuf Genc","submitted_at":"2024-11-21T11:20:48Z","abstract_excerpt":"We present two comprehensive benchmarks to evaluate the performance of language models in coding assistance tasks, covering code writing, debugging, code review, and conceptual understanding. Our main contribution includes two curated datasets: StackEval, a large-scale benchmark derived from Stack Overflow questions, and StackUnseen, a dynamic benchmark featuring the most recent Stack Overflow content. These benchmarks offer novel insights into the capabilities and limitations of LLMs, particularly in handling new and emerging content. Additionally, we assess LLMs' proficiency as judges for co"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.05288","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/2412.05288/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":"2026-07-05T09:46:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dAPWNKuAbpXzDhletq4Geezz3s5lsClDZDTTMIubJcHUfnXkBQrB4+h/gPXb4/Q3IVAU7b1muKhN7j18rZrAAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T03:04:42.349482Z"},"content_sha256":"6204fe5dd8e56f5e16dbffce6b951607f05b618fdf9f0a3ca7203d91b9d03288","schema_version":"1.0","event_id":"sha256:6204fe5dd8e56f5e16dbffce6b951607f05b618fdf9f0a3ca7203d91b9d03288"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/X7YDW5XK3DOYEAWIXIKAZEKWXR/bundle.json","state_url":"https://pith.science/pith/X7YDW5XK3DOYEAWIXIKAZEKWXR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/X7YDW5XK3DOYEAWIXIKAZEKWXR/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":"2026-08-18T03:04:42Z","links":{"resolver":"https://pith.science/pith/X7YDW5XK3DOYEAWIXIKAZEKWXR","bundle":"https://pith.science/pith/X7YDW5XK3DOYEAWIXIKAZEKWXR/bundle.json","state":"https://pith.science/pith/X7YDW5XK3DOYEAWIXIKAZEKWXR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/X7YDW5XK3DOYEAWIXIKAZEKWXR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:X7YDW5XK3DOYEAWIXIKAZEKWXR","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":"09a51479a820c58c959baa7b32b2f0b53c8e8b95311ff6ef43289f609af79908","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.SE","submitted_at":"2024-11-21T11:20:48Z","title_canon_sha256":"df53d77587772e4a4b8b1ba5ec018653539b98aacb7b8493e63b08ead4b64f06"},"schema_version":"1.0","source":{"id":"2412.05288","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.05288","created_at":"2026-07-05T09:46:00Z"},{"alias_kind":"arxiv_version","alias_value":"2412.05288v1","created_at":"2026-07-05T09:46:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.05288","created_at":"2026-07-05T09:46:00Z"},{"alias_kind":"pith_short_12","alias_value":"X7YDW5XK3DOY","created_at":"2026-07-05T09:46:00Z"},{"alias_kind":"pith_short_16","alias_value":"X7YDW5XK3DOYEAWI","created_at":"2026-07-05T09:46:00Z"},{"alias_kind":"pith_short_8","alias_value":"X7YDW5XK","created_at":"2026-07-05T09:46:00Z"}],"graph_snapshots":[{"event_id":"sha256:6204fe5dd8e56f5e16dbffce6b951607f05b618fdf9f0a3ca7203d91b9d03288","target":"graph","created_at":"2026-07-05T09:46:00Z","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/2412.05288/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We present two comprehensive benchmarks to evaluate the performance of language models in coding assistance tasks, covering code writing, debugging, code review, and conceptual understanding. Our main contribution includes two curated datasets: StackEval, a large-scale benchmark derived from Stack Overflow questions, and StackUnseen, a dynamic benchmark featuring the most recent Stack Overflow content. These benchmarks offer novel insights into the capabilities and limitations of LLMs, particularly in handling new and emerging content. Additionally, we assess LLMs' proficiency as judges for co","authors_text":"Dogu Araci, Nidhish Shah, Zulkuf Genc","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.SE","submitted_at":"2024-11-21T11:20:48Z","title":"StackEval: Benchmarking LLMs in Coding Assistance"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.05288","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":"sha256:704b310da832358772b04d38e70ae59793f4c99223fa621b603dafd9a77a2bfa","target":"record","created_at":"2026-07-05T09:46:00Z","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":"09a51479a820c58c959baa7b32b2f0b53c8e8b95311ff6ef43289f609af79908","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.SE","submitted_at":"2024-11-21T11:20:48Z","title_canon_sha256":"df53d77587772e4a4b8b1ba5ec018653539b98aacb7b8493e63b08ead4b64f06"},"schema_version":"1.0","source":{"id":"2412.05288","kind":"arxiv","version":1}},"canonical_sha256":"bff03b76ead8dd8202c8ba140c9156bc6568f38793f5e4b99ff82e47b1e2f4aa","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bff03b76ead8dd8202c8ba140c9156bc6568f38793f5e4b99ff82e47b1e2f4aa","first_computed_at":"2026-07-05T09:46:00.781147Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:46:00.781147Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"jcLbZveZZZdgMZqZ9MzTUyARl8uuk3tw4HTQJW298FCNONh1+qIBb0U1QsBRCx7yrx/7HTgSq0/0d15w6wxiAA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:46:00.781498Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.05288","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:704b310da832358772b04d38e70ae59793f4c99223fa621b603dafd9a77a2bfa","sha256:6204fe5dd8e56f5e16dbffce6b951607f05b618fdf9f0a3ca7203d91b9d03288"],"state_sha256":"cf6b942fdfce063551cd84882875667551dd01070cd80bc752e2cd828f9af43c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"j0WbFw+8MGCWjErA0pWs+5fUuO+L/fOkib1kAx4yAeBwXnp+W65RzcwxGkufz4jVT7DycPHLq9kADhBVnSY1DA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-18T03:04:42.354860Z","bundle_sha256":"853d3f68f7612e1a2ba447d48b02462fdc319ae3560934d10bc28de2c7a57c5a"}}