翻訳待ち：Detecting a Route Flip Is Easier Than Knowing Whether to Fix It: Causal Route-Mediated Damage in Quantized Mixture-of-Experts

AI サービスが一時的に利用できないため、復旧後に翻訳を補完します。

--> [Submitted on 14 Jul 2026] Title:Detecting a Route Flip Is Easier Than Knowing Whether to Fix It: Causal Route-Mediated Damage in Quantized Mixture-of-Experts View a PDF of the paper titled Detecting a Route Flip Is Easier Than Knowing Whether to Fix It: Causal Route-Mediated Damage in Quantized Mixture-of-Experts, by Parvel Gu View PDF HTML (experimental) Abstract:Top-k Mixture-of-Experts (MoE) routing is discontinuous, so a deployment-motivated numerical disturbance -- simulated 4-bit KV-cache quantization read by a protected BF16 gate -- pushes tokens across decision boundaries and flips which experts fire. This paper proposes no new mitigation; it supplies a causal apparatus, empirical findings, and a detection-limit result. A four-run apparatus prices the route-mediated fraction (RMF) of quantization damage, a token-level attribution decomposes it by mechanism, and pre-registered probes carry the findings across three architectures. On OLMoE-1B-7B at 4-bit KV (pilot), about a third of the damage is routing-mediated: RMF ~ 0.31 (discovery 0.31 [0.20, 0.41]; process-replicated mean 0.313 +/- 0.020; pre-registered re-execution 0.231). The deployable router margin detects that a flip occurred (AUC 0.772) but cannot tell a harmful flip from a helpful one (at chance): among the tested local, inference-observable router statistics we find no predictor of a flip's loss sign above chance -- an empirical benefit-detection barrier bounding selective repair restricted to this feature family. The signed-flip tax and sign-inseparability carry cross-model; the clean-reference remedy's payout is architecture-modulated; a controlled same-checkpoint flag-swap re-scopes the gate's normalization convention to a damage-magnitude moderator, not a route-recoverability mechanism. A real int4 KV kernel yields a fraction compatible with the fake-quant dose curve but underpowered (95% CI [-0.111, 0.394] includes zero) -- ruling out gross disagreement, not an independent replication. Hypotheses, thresholds, and evaluations were pre-registered before measurement, with misses reported; a pre-registered held-out read replicates the partition and the near-cancelling tax out of sample, while the strict impossibility exclusion narrowly misses. Comments: 13 pages, 2 figures, 8 tables. Pre-registered pilot study Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG) MSC classes: 68T07 ACM classes: I.2.6; I.2.7 Cite as: arXiv:2608.11212 [cs.AI] (or arXiv:2608.11212v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.11212 arXiv-issued DOI via DataCite Submission history From: Parvel Gu [view email] [v1] Tue, 14 Jul 2026 01:12:04 UTC (373 KB) Full-text links: Access Paper: View a PDF of the paper titled Detecting a Route Flip Is Easier Than Knowing Whether to Fix It: Causal Route-Mediated Damage in Quantized Mixture-of-Experts, by Parvel Gu View PDF HTML (experimental) TeX Source view license Ancillary-file links: Ancillary files (details): HASH_MANIFEST.txt PREREG_REGISTRY.md VERIFICATION.md prereg_keys/P1_jump_removable.txt prereg_keys/P2_kl_alignment_wa_not_kv.txt prereg_keys/P3_vanishing_boundary_rmf0.txt prereg_keys/P4_twin_gate.txt prereg_keys/P5_window_cusum_bound.txt prereg_keys/P8_crossover_generality.txt (4 additional files not shown) Current browse context: cs.AI new | recent | 2026-08 Change to browse by: cs cs.CL cs.LG References & Citations NASA ADS Google Scholar Semantic Scholar Loading... Data provided by: Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)