{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:NYQJW7YE4UJ4TEETYLPUAM4JYP","short_pith_number":"pith:NYQJW7YE","canonical_record":{"source":{"id":"2506.01205","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-01T22:50:06Z","cross_cats_sorted":[],"title_canon_sha256":"1a23ac1f1e014bc321832261368a5de946a9dd3c12532865486666bdbc2ed5b7","abstract_canon_sha256":"1ff07983e2516cd9a57a92a3360e257fefecd9a67e827dcad79528b206a40b30"},"schema_version":"1.0"},"canonical_sha256":"6e209b7f04e513c99093c2df403389c3c633811d38f679bedef386f3b5d38b96","source":{"kind":"arxiv","id":"2506.01205","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.01205","created_at":"2026-07-05T11:13:48Z"},{"alias_kind":"arxiv_version","alias_value":"2506.01205v1","created_at":"2026-07-05T11:13:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.01205","created_at":"2026-07-05T11:13:48Z"},{"alias_kind":"pith_short_12","alias_value":"NYQJW7YE4UJ4","created_at":"2026-07-05T11:13:48Z"},{"alias_kind":"pith_short_16","alias_value":"NYQJW7YE4UJ4TEET","created_at":"2026-07-05T11:13:48Z"},{"alias_kind":"pith_short_8","alias_value":"NYQJW7YE","created_at":"2026-07-05T11:13:48Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:NYQJW7YE4UJ4TEETYLPUAM4JYP","target":"record","payload":{"canonical_record":{"source":{"id":"2506.01205","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-01T22:50:06Z","cross_cats_sorted":[],"title_canon_sha256":"1a23ac1f1e014bc321832261368a5de946a9dd3c12532865486666bdbc2ed5b7","abstract_canon_sha256":"1ff07983e2516cd9a57a92a3360e257fefecd9a67e827dcad79528b206a40b30"},"schema_version":"1.0"},"canonical_sha256":"6e209b7f04e513c99093c2df403389c3c633811d38f679bedef386f3b5d38b96","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:13:48.068805Z","signature_b64":"hHbpiqkOmsHDrbHsXGffzzuxzo5ppvHk0WzNTQdkEF/71nS7be3nrn/58AjnyTygBwVZMM8uVG0JVXw5bsw9AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6e209b7f04e513c99093c2df403389c3c633811d38f679bedef386f3b5d38b96","last_reissued_at":"2026-07-05T11:13:48.068390Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:13:48.068390Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2506.01205","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":"2026-07-05T11:13:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"p0Dxw/XA6ulMLegFyB33yghMmpmX9USiQsuUpMYMmHtkBG1anynrKc7ku/hQMyaThPTwhB0UeVF16lzA2/foAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T22:51:54.594693Z"},"content_sha256":"61c29fbd0e743d1d1f2ff3d65b17b5f1f6706dac28b2522bf31ee1ebdcac21b9","schema_version":"1.0","event_id":"sha256:61c29fbd0e743d1d1f2ff3d65b17b5f1f6706dac28b2522bf31ee1ebdcac21b9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:NYQJW7YE4UJ4TEETYLPUAM4JYP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Trick or Neat: Adversarial Ambiguity and Language Model Evaluation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Anders S{\\o}gaard, Antonia Karamolegkou, Carina Kauf, Oliver Eberle, Phillip Rust","submitted_at":"2025-06-01T22:50:06Z","abstract_excerpt":"Detecting ambiguity is important for language understanding, including uncertainty estimation, humour detection, and processing garden path sentences. We assess language models' sensitivity to ambiguity by introducing an adversarial ambiguity dataset that includes syntactic, lexical, and phonological ambiguities along with adversarial variations (e.g., word-order changes, synonym replacements, and random-based alterations). Our findings show that direct prompting fails to robustly identify ambiguity, while linear probes trained on model representations can decode ambiguity with high accuracy, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.01205","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/2506.01205/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":"2026-07-05T11:13:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Kla56govUxyfX8/AbKulTXHqmdUbNipoLCNev7nQ+ciJAkHIrUvXGpJhf4Bjh9REevCjhqk6rNHFtQwRJn8qDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T22:51:54.595759Z"},"content_sha256":"d1fe2afdd9645ef9a6cec8c80f3d1153b70b84338742b97d66497a5695ee40ec","schema_version":"1.0","event_id":"sha256:d1fe2afdd9645ef9a6cec8c80f3d1153b70b84338742b97d66497a5695ee40ec"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/NYQJW7YE4UJ4TEETYLPUAM4JYP/bundle.json","state_url":"https://pith.science/pith/NYQJW7YE4UJ4TEETYLPUAM4JYP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/NYQJW7YE4UJ4TEETYLPUAM4JYP/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":"2026-08-17T22:51:54Z","links":{"resolver":"https://pith.science/pith/NYQJW7YE4UJ4TEETYLPUAM4JYP","bundle":"https://pith.science/pith/NYQJW7YE4UJ4TEETYLPUAM4JYP/bundle.json","state":"https://pith.science/pith/NYQJW7YE4UJ4TEETYLPUAM4JYP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/NYQJW7YE4UJ4TEETYLPUAM4JYP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:NYQJW7YE4UJ4TEETYLPUAM4JYP","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":"1ff07983e2516cd9a57a92a3360e257fefecd9a67e827dcad79528b206a40b30","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-01T22:50:06Z","title_canon_sha256":"1a23ac1f1e014bc321832261368a5de946a9dd3c12532865486666bdbc2ed5b7"},"schema_version":"1.0","source":{"id":"2506.01205","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.01205","created_at":"2026-07-05T11:13:48Z"},{"alias_kind":"arxiv_version","alias_value":"2506.01205v1","created_at":"2026-07-05T11:13:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.01205","created_at":"2026-07-05T11:13:48Z"},{"alias_kind":"pith_short_12","alias_value":"NYQJW7YE4UJ4","created_at":"2026-07-05T11:13:48Z"},{"alias_kind":"pith_short_16","alias_value":"NYQJW7YE4UJ4TEET","created_at":"2026-07-05T11:13:48Z"},{"alias_kind":"pith_short_8","alias_value":"NYQJW7YE","created_at":"2026-07-05T11:13:48Z"}],"graph_snapshots":[{"event_id":"sha256:d1fe2afdd9645ef9a6cec8c80f3d1153b70b84338742b97d66497a5695ee40ec","target":"graph","created_at":"2026-07-05T11:13:48Z","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/2506.01205/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Detecting ambiguity is important for language understanding, including uncertainty estimation, humour detection, and processing garden path sentences. We assess language models' sensitivity to ambiguity by introducing an adversarial ambiguity dataset that includes syntactic, lexical, and phonological ambiguities along with adversarial variations (e.g., word-order changes, synonym replacements, and random-based alterations). Our findings show that direct prompting fails to robustly identify ambiguity, while linear probes trained on model representations can decode ambiguity with high accuracy, ","authors_text":"Anders S{\\o}gaard, Antonia Karamolegkou, Carina Kauf, Oliver Eberle, Phillip Rust","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-01T22:50:06Z","title":"Trick or Neat: Adversarial Ambiguity and Language Model Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.01205","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":"sha256:61c29fbd0e743d1d1f2ff3d65b17b5f1f6706dac28b2522bf31ee1ebdcac21b9","target":"record","created_at":"2026-07-05T11:13:48Z","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":"1ff07983e2516cd9a57a92a3360e257fefecd9a67e827dcad79528b206a40b30","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-01T22:50:06Z","title_canon_sha256":"1a23ac1f1e014bc321832261368a5de946a9dd3c12532865486666bdbc2ed5b7"},"schema_version":"1.0","source":{"id":"2506.01205","kind":"arxiv","version":1}},"canonical_sha256":"6e209b7f04e513c99093c2df403389c3c633811d38f679bedef386f3b5d38b96","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6e209b7f04e513c99093c2df403389c3c633811d38f679bedef386f3b5d38b96","first_computed_at":"2026-07-05T11:13:48.068390Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:13:48.068390Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hHbpiqkOmsHDrbHsXGffzzuxzo5ppvHk0WzNTQdkEF/71nS7be3nrn/58AjnyTygBwVZMM8uVG0JVXw5bsw9AA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:13:48.068805Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.01205","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:61c29fbd0e743d1d1f2ff3d65b17b5f1f6706dac28b2522bf31ee1ebdcac21b9","sha256:d1fe2afdd9645ef9a6cec8c80f3d1153b70b84338742b97d66497a5695ee40ec"],"state_sha256":"a645fca48ab8a334661e70006cd5d401cb4c86d141e5df5d8caa3c8be2e97633"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mKAsJBNFVQiYfWS5+BuTofqh2686oLuZcI2R8M2vQj79vThp8N9LfaQw64qw4KjO8+hAW9caf4wJiipQG2xAAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-17T22:51:54.602547Z","bundle_sha256":"b3d91ac730ce227aa74c41847c0615594af45a3079ba986f0dd83411bc14a8c5"}}