{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:BDDQ3SU5C4FNMMBT2LSZ5F54N4","short_pith_number":"pith:BDDQ3SU5","canonical_record":{"source":{"id":"2605.12620","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-12T18:08:24Z","cross_cats_sorted":[],"title_canon_sha256":"df1ac8fdff8401667f177b8bd831d1159e04ca4b32cb55a1738c26c46f7d3af5","abstract_canon_sha256":"357ae961b7ad0cd33cf52aba688de6cfeb3ccedb2b2a1e58e3e3dc61b353b0ae"},"schema_version":"1.0"},"canonical_sha256":"08c70dca9d170ad63033d2e59e97bc6f28bb589f8e93b171468152d78bb4ad37","source":{"kind":"arxiv","id":"2605.12620","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.12620","created_at":"2026-05-18T03:10:00Z"},{"alias_kind":"arxiv_version","alias_value":"2605.12620v1","created_at":"2026-05-18T03:10:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.12620","created_at":"2026-05-18T03:10:00Z"},{"alias_kind":"pith_short_12","alias_value":"BDDQ3SU5C4FN","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"BDDQ3SU5C4FNMMBT","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"BDDQ3SU5","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:BDDQ3SU5C4FNMMBT2LSZ5F54N4","target":"record","payload":{"canonical_record":{"source":{"id":"2605.12620","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-12T18:08:24Z","cross_cats_sorted":[],"title_canon_sha256":"df1ac8fdff8401667f177b8bd831d1159e04ca4b32cb55a1738c26c46f7d3af5","abstract_canon_sha256":"357ae961b7ad0cd33cf52aba688de6cfeb3ccedb2b2a1e58e3e3dc61b353b0ae"},"schema_version":"1.0"},"canonical_sha256":"08c70dca9d170ad63033d2e59e97bc6f28bb589f8e93b171468152d78bb4ad37","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T03:10:00.448101Z","signature_b64":"IN4Pv1TQibiSC+bPRIhs+huFCwXHepPs1IfjtYXeW+ZQf0QhGULhWWMa1SGkvGG4ofqR3XBVFpXBx0y/OghADQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"08c70dca9d170ad63033d2e59e97bc6f28bb589f8e93b171468152d78bb4ad37","last_reissued_at":"2026-05-18T03:10:00.447469Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T03:10:00.447469Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.12620","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":"2026-05-18T03:10:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KNNoS7jvqSBLa0pfLOldJBXxqS/PGjzNfCtfLKPUlW0ELA+n3oRA0WQKPby9nA1AGBGOtWc4yNWW06QaWqklBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-19T20:11:37.328842Z"},"content_sha256":"1c290631c4a03047a1d321b2d20e749b67efbef6ca74fe035c54f9b42940d0dd","schema_version":"1.0","event_id":"sha256:1c290631c4a03047a1d321b2d20e749b67efbef6ca74fe035c54f9b42940d0dd"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:BDDQ3SU5C4FNMMBT2LSZ5F54N4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Think Twice, Act Once: Verifier-Guided Action Selection For Embodied Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A test-time verifier trained on synthesized failures helps MLLM agents pick reliable actions from multiple candidates.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Anna Rohrbach, Christian Bialas, Georgia Chalvatzaki, Marcus Rohrbach, Nishad Singhi, Snehal Jauhri, Vignesh Prasad","submitted_at":"2026-05-12T18:08:24Z","abstract_excerpt":"Building generalist embodied agents capable of solving complex real-world tasks remains a fundamental challenge in AI. Multimodal Large Language Models (MLLMs) have significantly advanced the reasoning capabilities of such agents through strong vision-language knowledge and chain-of-thought (CoT) reasoning, yet remain brittle when faced with challenging out-of-distribution scenarios. To address this, we propose Verifier-Guided Action Selection (VegAS), a test-time framework designed to improve the robustness of MLLM-based embodied agents through an explicit verification step. At inference time"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Across embodied reasoning benchmarks spanning the Habitat and ALFRED environments, VeGAS consistently improves generalization, achieving up to a 36% relative performance gain over strong CoT baselines on the most challenging multi-object, long-horizon tasks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That training a verifier on automatically synthesized failure cases from an LLM will produce a model that reliably identifies good actions in out-of-distribution scenarios where the base MLLM fails.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"VeGAS improves MLLM-based embodied agents by sampling action ensembles and using a verifier trained on LLM-synthesized failure cases, yielding up to 36% relative gains on hard multi-object long-horizon tasks in Habitat and ALFRED.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A test-time verifier trained on synthesized failures helps MLLM agents pick reliable actions from multiple candidates.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"9e87c00acfd66d34ca16915d4712344a09642c67cd415fa10b35d95b88e72510"},"source":{"id":"2605.12620","kind":"arxiv","version":1},"verdict":{"id":"1e46b4c7-8745-451b-91f1-81cf25d3ab8e","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T21:00:48.159839Z","strongest_claim":"Across embodied reasoning benchmarks spanning the Habitat and ALFRED environments, VeGAS consistently improves generalization, achieving up to a 36% relative performance gain over strong CoT baselines on the most challenging multi-object, long-horizon tasks.","one_line_summary":"VeGAS improves MLLM-based embodied agents by sampling action ensembles and using a verifier trained on LLM-synthesized failure cases, yielding up to 36% relative gains on hard multi-object long-horizon tasks in Habitat and ALFRED.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That training a verifier on automatically synthesized failure cases from an LLM will produce a model that reliably identifies good actions in out-of-distribution scenarios where the base MLLM fails.","pith_extraction_headline":"A test-time verifier trained on synthesized failures helps MLLM agents pick reliable actions from multiple candidates."},"references":{"count":65,"sample":[{"doi":"","year":2022,"title":"Do As I Can, Not As I Say: Grounding Language in Robotic Affordances","work_id":"037320f1-b0a9-4cbe-a639-bfb25409ce71","ref_index":1,"cited_arxiv_id":"2204.01691","is_internal_anchor":true},{"doi":"","year":2024,"title":"Critique-out-loud reward models","work_id":"adbb5d79-de43-49c4-abd5-ad004e77abea","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Qwen2.5-VL Technical Report","work_id":"69dffacb-bfe8-442d-be86-48624c60426f","ref_index":3,"cited_arxiv_id":"2502.13923","is_internal_anchor":true},{"doi":"","year":2024,"title":"Large Language Monkeys: Scaling Inference Compute with Repeated Sampling","work_id":"b124064d-5a56-42ad-86f5-3cc349b86a3a","ref_index":4,"cited_arxiv_id":"2407.21787","is_internal_anchor":true},{"doi":"","year":2021,"title":"Training Verifiers to Solve Math Word Problems","work_id":"acab1aa8-b4d6-40e0-a3ee-25341701dca2","ref_index":5,"cited_arxiv_id":"2110.14168","is_internal_anchor":true}],"resolved_work":65,"snapshot_sha256":"c641e77b14dc291d581e8353a882323638234b3b5972594db6be58da7cac2ea4","internal_anchors":7},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ccccfa6436f3acdb87894e5442821afb4e80ad60b5f5b3ed0c56ec6090bc8829"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"1e46b4c7-8745-451b-91f1-81cf25d3ab8e"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T03:10:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Je/IRmfwYJPJqFnEwl86VLUWw8rrkBHd64FiC2J7X9KarMGkgu4AMzSoaO2n1MREIDDsSq8Msk9g+VQqp3PHDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-19T20:11:37.329380Z"},"content_sha256":"1994b57c182e1aed7a8f2bbf1aa59a6c28a5560eb8806c26e1ef7204c6b44c0f","schema_version":"1.0","event_id":"sha256:1994b57c182e1aed7a8f2bbf1aa59a6c28a5560eb8806c26e1ef7204c6b44c0f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/BDDQ3SU5C4FNMMBT2LSZ5F54N4/bundle.json","state_url":"https://pith.science/pith/BDDQ3SU5C4FNMMBT2LSZ5F54N4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/BDDQ3SU5C4FNMMBT2LSZ5F54N4/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":"2026-05-19T20:11:37Z","links":{"resolver":"https://pith.science/pith/BDDQ3SU5C4FNMMBT2LSZ5F54N4","bundle":"https://pith.science/pith/BDDQ3SU5C4FNMMBT2LSZ5F54N4/bundle.json","state":"https://pith.science/pith/BDDQ3SU5C4FNMMBT2LSZ5F54N4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/BDDQ3SU5C4FNMMBT2LSZ5F54N4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:BDDQ3SU5C4FNMMBT2LSZ5F54N4","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":"357ae961b7ad0cd33cf52aba688de6cfeb3ccedb2b2a1e58e3e3dc61b353b0ae","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-12T18:08:24Z","title_canon_sha256":"df1ac8fdff8401667f177b8bd831d1159e04ca4b32cb55a1738c26c46f7d3af5"},"schema_version":"1.0","source":{"id":"2605.12620","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.12620","created_at":"2026-05-18T03:10:00Z"},{"alias_kind":"arxiv_version","alias_value":"2605.12620v1","created_at":"2026-05-18T03:10:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.12620","created_at":"2026-05-18T03:10:00Z"},{"alias_kind":"pith_short_12","alias_value":"BDDQ3SU5C4FN","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"BDDQ3SU5C4FNMMBT","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"BDDQ3SU5","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:1994b57c182e1aed7a8f2bbf1aa59a6c28a5560eb8806c26e1ef7204c6b44c0f","target":"graph","created_at":"2026-05-18T03:10:00Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Across embodied reasoning benchmarks spanning the Habitat and ALFRED environments, VeGAS consistently improves generalization, achieving up to a 36% relative performance gain over strong CoT baselines on the most challenging multi-object, long-horizon tasks."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That training a verifier on automatically synthesized failure cases from an LLM will produce a model that reliably identifies good actions in out-of-distribution scenarios where the base MLLM fails."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"VeGAS improves MLLM-based embodied agents by sampling action ensembles and using a verifier trained on LLM-synthesized failure cases, yielding up to 36% relative gains on hard multi-object long-horizon tasks in Habitat and ALFRED."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A test-time verifier trained on synthesized failures helps MLLM agents pick reliable actions from multiple candidates."}],"snapshot_sha256":"9e87c00acfd66d34ca16915d4712344a09642c67cd415fa10b35d95b88e72510"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ccccfa6436f3acdb87894e5442821afb4e80ad60b5f5b3ed0c56ec6090bc8829"},"paper":{"abstract_excerpt":"Building generalist embodied agents capable of solving complex real-world tasks remains a fundamental challenge in AI. Multimodal Large Language Models (MLLMs) have significantly advanced the reasoning capabilities of such agents through strong vision-language knowledge and chain-of-thought (CoT) reasoning, yet remain brittle when faced with challenging out-of-distribution scenarios. To address this, we propose Verifier-Guided Action Selection (VegAS), a test-time framework designed to improve the robustness of MLLM-based embodied agents through an explicit verification step. At inference time","authors_text":"Anna Rohrbach, Christian Bialas, Georgia Chalvatzaki, Marcus Rohrbach, Nishad Singhi, Snehal Jauhri, Vignesh Prasad","cross_cats":[],"headline":"A test-time verifier trained on synthesized failures helps MLLM agents pick reliable actions from multiple candidates.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-12T18:08:24Z","title":"Think Twice, Act Once: Verifier-Guided Action Selection For Embodied Agents"},"references":{"count":65,"internal_anchors":7,"resolved_work":65,"sample":[{"cited_arxiv_id":"2204.01691","doi":"","is_internal_anchor":true,"ref_index":1,"title":"Do As I Can, Not As I Say: Grounding Language in Robotic Affordances","work_id":"037320f1-b0a9-4cbe-a639-bfb25409ce71","year":2022},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Critique-out-loud reward models","work_id":"adbb5d79-de43-49c4-abd5-ad004e77abea","year":2024},{"cited_arxiv_id":"2502.13923","doi":"","is_internal_anchor":true,"ref_index":3,"title":"Qwen2.5-VL Technical Report","work_id":"69dffacb-bfe8-442d-be86-48624c60426f","year":null},{"cited_arxiv_id":"2407.21787","doi":"","is_internal_anchor":true,"ref_index":4,"title":"Large Language Monkeys: Scaling Inference Compute with Repeated Sampling","work_id":"b124064d-5a56-42ad-86f5-3cc349b86a3a","year":2024},{"cited_arxiv_id":"2110.14168","doi":"","is_internal_anchor":true,"ref_index":5,"title":"Training Verifiers to Solve Math Word Problems","work_id":"acab1aa8-b4d6-40e0-a3ee-25341701dca2","year":2021}],"snapshot_sha256":"c641e77b14dc291d581e8353a882323638234b3b5972594db6be58da7cac2ea4"},"source":{"id":"2605.12620","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T21:00:48.159839Z","id":"1e46b4c7-8745-451b-91f1-81cf25d3ab8e","model_set":{"reader":"grok-4.3"},"one_line_summary":"VeGAS improves MLLM-based embodied agents by sampling action ensembles and using a verifier trained on LLM-synthesized failure cases, yielding up to 36% relative gains on hard multi-object long-horizon tasks in Habitat and ALFRED.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A test-time verifier trained on synthesized failures helps MLLM agents pick reliable actions from multiple candidates.","strongest_claim":"Across embodied reasoning benchmarks spanning the Habitat and ALFRED environments, VeGAS consistently improves generalization, achieving up to a 36% relative performance gain over strong CoT baselines on the most challenging multi-object, long-horizon tasks.","weakest_assumption":"That training a verifier on automatically synthesized failure cases from an LLM will produce a model that reliably identifies good actions in out-of-distribution scenarios where the base MLLM fails."}},"verdict_id":"1e46b4c7-8745-451b-91f1-81cf25d3ab8e"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:1c290631c4a03047a1d321b2d20e749b67efbef6ca74fe035c54f9b42940d0dd","target":"record","created_at":"2026-05-18T03:10:00Z","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":"357ae961b7ad0cd33cf52aba688de6cfeb3ccedb2b2a1e58e3e3dc61b353b0ae","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-12T18:08:24Z","title_canon_sha256":"df1ac8fdff8401667f177b8bd831d1159e04ca4b32cb55a1738c26c46f7d3af5"},"schema_version":"1.0","source":{"id":"2605.12620","kind":"arxiv","version":1}},"canonical_sha256":"08c70dca9d170ad63033d2e59e97bc6f28bb589f8e93b171468152d78bb4ad37","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"08c70dca9d170ad63033d2e59e97bc6f28bb589f8e93b171468152d78bb4ad37","first_computed_at":"2026-05-18T03:10:00.447469Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T03:10:00.447469Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"IN4Pv1TQibiSC+bPRIhs+huFCwXHepPs1IfjtYXeW+ZQf0QhGULhWWMa1SGkvGG4ofqR3XBVFpXBx0y/OghADQ==","signature_status":"signed_v1","signed_at":"2026-05-18T03:10:00.448101Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.12620","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1c290631c4a03047a1d321b2d20e749b67efbef6ca74fe035c54f9b42940d0dd","sha256:1994b57c182e1aed7a8f2bbf1aa59a6c28a5560eb8806c26e1ef7204c6b44c0f"],"state_sha256":"405d2d42516d93fde28e48ae54f6da274e11e218349612b2c73ea7afcd11029d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"f+xlV8/ZzFYO5jo8PZGwkCCAuwbucRG8pNju2XAu/KuP0seJLGqtyP4YnuBQRT70RGusChJCGNIlGR63wc09Ag==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-19T20:11:37.333200Z","bundle_sha256":"db79c5e36a15ff7bc1d737ba025e53dbcae637ca9aeb80a783f1917d4f65bf54"}}