{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:QVISCZETLXX7NKDPOS6N532WAN","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":"ba16023a18f482c3f4e89c9c73a902f8fcfe2dffd74f07a9403b1c06e1db13b3","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-07T18:34:12Z","title_canon_sha256":"8b6ad8640fe227bb5b4ecb2c6c5e6519aaf86bf0bc307af4e6be6dce01563014"},"schema_version":"1.0","source":{"id":"2605.06831","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.06831","created_at":"2026-06-01T01:02:41Z"},{"alias_kind":"arxiv_version","alias_value":"2605.06831v2","created_at":"2026-06-01T01:02:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.06831","created_at":"2026-06-01T01:02:41Z"},{"alias_kind":"pith_short_12","alias_value":"QVISCZETLXX7","created_at":"2026-06-01T01:02:41Z"},{"alias_kind":"pith_short_16","alias_value":"QVISCZETLXX7NKDP","created_at":"2026-06-01T01:02:41Z"},{"alias_kind":"pith_short_8","alias_value":"QVISCZET","created_at":"2026-06-01T01:02:41Z"}],"graph_snapshots":[{"event_id":"sha256:2e5bbb3d5eb46cbc6ab78b74baa4cbe6dc5133c28c1e27c7a1f4e6df4c26746f","target":"graph","created_at":"2026-06-01T01:02:41Z","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":"We analyze the reverse ODE (DDIM) and SDE (DDPM) for a Gaussian mixture target, proving that after a critical time τ, (a) DDIM can become stuck on the segment connecting the two nearest modes and (b) DDPM *stochasticity* helps it become unstuck from this region, thus avoiding hallucination."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The theoretical results and empirical observations are derived for a low-dimensional Gaussian mixture target; the same trapping behavior and benefit of stochasticity may not occur or may be harder to characterize for high-dimensional, non-Gaussian data distributions used in practice."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"DDIM gets stuck between modes in reverse diffusion on Gaussian mixtures after critical time τ, but DDPM stochasticity prevents this and lowers hallucination rates."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"DDIM reverse trajectories can trap on the line between two modes after a critical time, while DDPM noise lets them escape and reach the true modes."}],"snapshot_sha256":"5ae3a7f4ea2d47321148418e583b19531664d21c8ead6de4fa64486b0c427905"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ba5d9505a2e2deb383cf8e71eae5fca22c28c51f84e4e3254d9ad4f2de41b206"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"claim_evidence","ran_at":"2026-05-20T12:02:03.778189Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-20T07:34:09.925892Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-19T18:01:18.715433Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T12:22:34.822230Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2605.06831/integrity.json","findings":[],"snapshot_sha256":"0e3735f6ae796db1a6957894555c367fc64a16be58a0e7b62c181d3d345a0023","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We theoretically study the hallucination phenomena in two canonical diffusion samplers: the stochastic Denoising Diffusion Probabilistic Model (DDPM) and the deterministic Denoising Diffusion Implicit Model (DDIM). We analyze the reverse ODE (DDIM) and SDE (DDPM) for a Gaussian mixture target, proving that after a critical time $\\tau$, (a) DDIM can become stuck on the segment connecting the two nearest modes and (b) DDPM *stochasticity* helps it become unstuck from this region, thus avoiding hallucination. Our empirical validation verifies that DDPM has a significantly lower hallucination rate","authors_text":"Abhinav N. Harish, Grigorios G. Chrysos, Hung Yun Tseng, Ishaan Kharbanda, Muhammad H. Ashiq, Samanyu Arora","cross_cats":["cs.AI"],"headline":"DDIM reverse trajectories can trap on the line between two modes after a critical time, while DDPM noise lets them escape and reach the true modes.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-07T18:34:12Z","title":"Why DDIM Hallucinates More Than DDPM: A Theoretical Analysis of Reverse Dynamics"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.06831","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-11T00:55:51.480715Z","id":"5bca75f3-6297-4762-b033-2977adb5dc27","model_set":{"reader":"grok-4.3"},"one_line_summary":"DDIM gets stuck between modes in reverse diffusion on Gaussian mixtures after critical time τ, but DDPM stochasticity prevents this and lowers hallucination rates.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"DDIM reverse trajectories can trap on the line between two modes after a critical time, while DDPM noise lets them escape and reach the true modes.","strongest_claim":"We analyze the reverse ODE (DDIM) and SDE (DDPM) for a Gaussian mixture target, proving that after a critical time τ, (a) DDIM can become stuck on the segment connecting the two nearest modes and (b) DDPM *stochasticity* helps it become unstuck from this region, thus avoiding hallucination.","weakest_assumption":"The theoretical results and empirical observations are derived for a low-dimensional Gaussian mixture target; the same trapping behavior and benefit of stochasticity may not occur or may be harder to characterize for high-dimensional, non-Gaussian data distributions used in practice."}},"verdict_id":"5bca75f3-6297-4762-b033-2977adb5dc27"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:c4dc3eafa8bdda0558445d8731009ab84503bfb34900d0bb16b45739cccaa4d3","target":"record","created_at":"2026-06-01T01:02:41Z","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":"ba16023a18f482c3f4e89c9c73a902f8fcfe2dffd74f07a9403b1c06e1db13b3","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-07T18:34:12Z","title_canon_sha256":"8b6ad8640fe227bb5b4ecb2c6c5e6519aaf86bf0bc307af4e6be6dce01563014"},"schema_version":"1.0","source":{"id":"2605.06831","kind":"arxiv","version":2}},"canonical_sha256":"85512164935deff6a86f74bcdeef5603420cc74837beef9119d33065ed6598e6","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"85512164935deff6a86f74bcdeef5603420cc74837beef9119d33065ed6598e6","first_computed_at":"2026-06-01T01:02:41.899643Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-01T01:02:41.899643Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"0pIpZESQ5J294aWsQ4VqhNQL6zqDPD0v8fXrlE0QV/CWnBx6q7B5c9uJgSRqEvVfhwp2kPscg1oQVtPg5KUeAA==","signature_status":"signed_v1","signed_at":"2026-06-01T01:02:41.900635Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.06831","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c4dc3eafa8bdda0558445d8731009ab84503bfb34900d0bb16b45739cccaa4d3","sha256:2e5bbb3d5eb46cbc6ab78b74baa4cbe6dc5133c28c1e27c7a1f4e6df4c26746f"],"state_sha256":"610ea59e61c506cf0655e8ff0a345c55e732af67f0f610eeb9dca40e41db93a4"}