{"paper":{"title":"Drift-AR: Single-Step Visual Autoregressive Generation via Anti-Symmetric Drifting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Drift-AR uses per-position prediction entropy to drive both speculative AR drafting and anti-symmetric drift, achieving genuine single-step visual generation.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Feng Zhao, Jie Huang, LinJiang Huang, Mingde Yao, Xiaoxiao Ma, Zhen Zou","submitted_at":"2026-03-30T05:29:00Z","abstract_excerpt":"Autoregressive (AR)-Diffusion hybrid paradigms combine AR's structured semantic modeling with diffusion's high-fidelity synthesis, yet suffer from a dual speed bottleneck: the sequential AR stage and the iterative multi-step denoising of the diffusion vision decode stage. Existing methods address each in isolation without a unified principle design. We observe that the per-position \\emph{prediction entropy} of continuous-space AR models naturally encodes spatially varying generation uncertainty, which simultaneously governing draft prediction quality in the AR stage and reflecting the correcti"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Drift-AR leverages entropy signal to accelerate both stages... enabling single-step (1-NFE) decoding without iterative denoising or distillation... Experiments on MAR, TransDiff, and NextStep-1 demonstrate 3.8-5.5× speedup with genuine 1-NFE decoding, matching or surpassing original quality.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The per-position prediction entropy of continuous-space AR models naturally encodes spatially varying generation uncertainty, which simultaneously governing draft prediction quality in the AR stage and reflecting the corrective effort required by vision decoding stage.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Drift-AR achieves 3.8-5.5x speedup in AR-diffusion image models by using entropy to enable entropy-informed speculative decoding and single-step (1-NFE) anti-symmetric drifting decoding.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Drift-AR uses per-position prediction entropy to drive both speculative AR drafting and anti-symmetric drift, achieving genuine single-step visual generation.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"dd803163490247bd8f743db9c349ae0fbb8866672a42e0320a03197a4a21a3aa"},"source":{"id":"2603.28049","kind":"arxiv","version":3},"verdict":{"id":"4903fae9-2e79-4fc6-bbe1-0a0da6a6f187","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T21:59:56.704758Z","strongest_claim":"Drift-AR leverages entropy signal to accelerate both stages... enabling single-step (1-NFE) decoding without iterative denoising or distillation... Experiments on MAR, TransDiff, and NextStep-1 demonstrate 3.8-5.5× speedup with genuine 1-NFE decoding, matching or surpassing original quality.","one_line_summary":"Drift-AR achieves 3.8-5.5x speedup in AR-diffusion image models by using entropy to enable entropy-informed speculative decoding and single-step (1-NFE) anti-symmetric drifting decoding.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The per-position prediction entropy of continuous-space AR models naturally encodes spatially varying generation uncertainty, which simultaneously governing draft prediction quality in the AR stage and reflecting the corrective effort required by vision decoding stage.","pith_extraction_headline":"Drift-AR uses per-position prediction entropy to drive both speculative AR drafting and anti-symmetric drift, achieving genuine single-step visual generation."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2603.28049/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"}