{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:3C2YUL3WGOXEIKSTXBJUU32LI7","short_pith_number":"pith:3C2YUL3W","canonical_record":{"source":{"id":"2406.12814","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-18T17:32:48Z","cross_cats_sorted":["cs.CL","cs.CR","cs.CV"],"title_canon_sha256":"f50434649805a594a7557a7f0029b03e93c75f9a95e7e276fee0068faaa671a2","abstract_canon_sha256":"33cbe7cf2eee97efa438bc7b11d9c133904bd2bb91798fea854620595360d8a0"},"schema_version":"1.0"},"canonical_sha256":"d8b58a2f7633ae442a53b8534a6f4b47c4c5d5bb4437936eff22e1c4fe33a27d","source":{"kind":"arxiv","id":"2406.12814","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.12814","created_at":"2026-07-05T10:09:44Z"},{"alias_kind":"arxiv_version","alias_value":"2406.12814v3","created_at":"2026-07-05T10:09:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.12814","created_at":"2026-07-05T10:09:44Z"},{"alias_kind":"pith_short_12","alias_value":"3C2YUL3WGOXE","created_at":"2026-07-05T10:09:44Z"},{"alias_kind":"pith_short_16","alias_value":"3C2YUL3WGOXEIKST","created_at":"2026-07-05T10:09:44Z"},{"alias_kind":"pith_short_8","alias_value":"3C2YUL3W","created_at":"2026-07-05T10:09:44Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:3C2YUL3WGOXEIKSTXBJUU32LI7","target":"record","payload":{"canonical_record":{"source":{"id":"2406.12814","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-18T17:32:48Z","cross_cats_sorted":["cs.CL","cs.CR","cs.CV"],"title_canon_sha256":"f50434649805a594a7557a7f0029b03e93c75f9a95e7e276fee0068faaa671a2","abstract_canon_sha256":"33cbe7cf2eee97efa438bc7b11d9c133904bd2bb91798fea854620595360d8a0"},"schema_version":"1.0"},"canonical_sha256":"d8b58a2f7633ae442a53b8534a6f4b47c4c5d5bb4437936eff22e1c4fe33a27d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:09:44.936860Z","signature_b64":"DpNRho4ZDJ/yYcnH4HW/a/l4ghyMFkaqt7qa079rHW2lqlhRBRHpeegBqphWAURYJAV9XhEeQLEpN2uGz/mkAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d8b58a2f7633ae442a53b8534a6f4b47c4c5d5bb4437936eff22e1c4fe33a27d","last_reissued_at":"2026-07-05T10:09:44.936382Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:09:44.936382Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.12814","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:09:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GHx2MniZDOdhTk8q2YbuIahzh4FI3WsG6lEF/4t9UxjlYjZAXjGbEdKhbCedPJ+Ld8CcERkoPCo0H6gm1GZGBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T06:52:23.282416Z"},"content_sha256":"dd3c1dc297a49d014a01cf849e696998f233750cadffd76fe35b851628ee0aac","schema_version":"1.0","event_id":"sha256:dd3c1dc297a49d014a01cf849e696998f233750cadffd76fe35b851628ee0aac"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:3C2YUL3WGOXEIKSTXBJUU32LI7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Dissecting Adversarial Robustness of Multimodal LM Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.CR","cs.CV"],"primary_cat":"cs.LG","authors_text":"Aditi Raghunathan, Chen Henry Wu, Daniel Fried, Jing Yu Koh, Rishi Shah, Ruslan Salakhutdinov","submitted_at":"2024-06-18T17:32:48Z","abstract_excerpt":"As language models (LMs) are used to build autonomous agents in real environments, ensuring their adversarial robustness becomes a critical challenge. Unlike chatbots, agents are compound systems with multiple components taking actions, which existing LMs safety evaluations do not adequately address. To bridge this gap, we manually create 200 targeted adversarial tasks and evaluation scripts in a realistic threat model on top of VisualWebArena, a real environment for web agents. To systematically examine the robustness of agents, we propose the Agent Robustness Evaluation (ARE) framework. ARE "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.12814","kind":"arxiv","version":3},"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/2406.12814/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":"2026-07-05T10:09:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"o6b/Emaga3kPxpHBFKHScHoR6jxI1Lt5nki1kLpNRhYBWgLQkguloHd9mWP8mv8Gjc7LR5MzO8AJu3k0m/8IDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T06:52:23.282800Z"},"content_sha256":"874c19fff14920905b312a7030270019495d694f2033380aa3eb42478cde00fc","schema_version":"1.0","event_id":"sha256:874c19fff14920905b312a7030270019495d694f2033380aa3eb42478cde00fc"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3C2YUL3WGOXEIKSTXBJUU32LI7/bundle.json","state_url":"https://pith.science/pith/3C2YUL3WGOXEIKSTXBJUU32LI7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3C2YUL3WGOXEIKSTXBJUU32LI7/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":"2026-08-04T06:52:23Z","links":{"resolver":"https://pith.science/pith/3C2YUL3WGOXEIKSTXBJUU32LI7","bundle":"https://pith.science/pith/3C2YUL3WGOXEIKSTXBJUU32LI7/bundle.json","state":"https://pith.science/pith/3C2YUL3WGOXEIKSTXBJUU32LI7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3C2YUL3WGOXEIKSTXBJUU32LI7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:3C2YUL3WGOXEIKSTXBJUU32LI7","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":"33cbe7cf2eee97efa438bc7b11d9c133904bd2bb91798fea854620595360d8a0","cross_cats_sorted":["cs.CL","cs.CR","cs.CV"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-18T17:32:48Z","title_canon_sha256":"f50434649805a594a7557a7f0029b03e93c75f9a95e7e276fee0068faaa671a2"},"schema_version":"1.0","source":{"id":"2406.12814","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.12814","created_at":"2026-07-05T10:09:44Z"},{"alias_kind":"arxiv_version","alias_value":"2406.12814v3","created_at":"2026-07-05T10:09:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.12814","created_at":"2026-07-05T10:09:44Z"},{"alias_kind":"pith_short_12","alias_value":"3C2YUL3WGOXE","created_at":"2026-07-05T10:09:44Z"},{"alias_kind":"pith_short_16","alias_value":"3C2YUL3WGOXEIKST","created_at":"2026-07-05T10:09:44Z"},{"alias_kind":"pith_short_8","alias_value":"3C2YUL3W","created_at":"2026-07-05T10:09:44Z"}],"graph_snapshots":[{"event_id":"sha256:874c19fff14920905b312a7030270019495d694f2033380aa3eb42478cde00fc","target":"graph","created_at":"2026-07-05T10:09:44Z","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/2406.12814/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As language models (LMs) are used to build autonomous agents in real environments, ensuring their adversarial robustness becomes a critical challenge. Unlike chatbots, agents are compound systems with multiple components taking actions, which existing LMs safety evaluations do not adequately address. To bridge this gap, we manually create 200 targeted adversarial tasks and evaluation scripts in a realistic threat model on top of VisualWebArena, a real environment for web agents. To systematically examine the robustness of agents, we propose the Agent Robustness Evaluation (ARE) framework. ARE ","authors_text":"Aditi Raghunathan, Chen Henry Wu, Daniel Fried, Jing Yu Koh, Rishi Shah, Ruslan Salakhutdinov","cross_cats":["cs.CL","cs.CR","cs.CV"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-18T17:32:48Z","title":"Dissecting Adversarial Robustness of Multimodal LM Agents"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.12814","kind":"arxiv","version":3},"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:dd3c1dc297a49d014a01cf849e696998f233750cadffd76fe35b851628ee0aac","target":"record","created_at":"2026-07-05T10:09:44Z","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":"33cbe7cf2eee97efa438bc7b11d9c133904bd2bb91798fea854620595360d8a0","cross_cats_sorted":["cs.CL","cs.CR","cs.CV"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-18T17:32:48Z","title_canon_sha256":"f50434649805a594a7557a7f0029b03e93c75f9a95e7e276fee0068faaa671a2"},"schema_version":"1.0","source":{"id":"2406.12814","kind":"arxiv","version":3}},"canonical_sha256":"d8b58a2f7633ae442a53b8534a6f4b47c4c5d5bb4437936eff22e1c4fe33a27d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d8b58a2f7633ae442a53b8534a6f4b47c4c5d5bb4437936eff22e1c4fe33a27d","first_computed_at":"2026-07-05T10:09:44.936382Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:09:44.936382Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"DpNRho4ZDJ/yYcnH4HW/a/l4ghyMFkaqt7qa079rHW2lqlhRBRHpeegBqphWAURYJAV9XhEeQLEpN2uGz/mkAA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:09:44.936860Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.12814","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:dd3c1dc297a49d014a01cf849e696998f233750cadffd76fe35b851628ee0aac","sha256:874c19fff14920905b312a7030270019495d694f2033380aa3eb42478cde00fc"],"state_sha256":"bb11934c0c34ffd4a2ce8e0a867f6d2acc54a346d766e499055143b7dd4486b6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tdf55Ql3faD/aRJZJciwxDyltZuYI3EIItRRuvpbzQe9UWaD/NvEyN6zcCQ+efFB710Lp+Rg3WQqGddlahTWBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T06:52:23.285263Z","bundle_sha256":"fb76b9398503f4350bb14a61aa924c4313c1e426e8714c1d7a0899d9e03007c3"}}