← 研踪 PaperTrailarXiv:2611.15513 · v1
When Cause Attribution Is Not Enough: Action-Value Control for Task-Oriented Semantic Communication with Discrete Encoder Switching
首次提交 2026-11-11 22:07
版本更新 2026-11-11 22:07
雷达观测 2026-11-13 11:11
摘要
A task-oriented semantic communication receiver that selects among independently trained encoders must answer two questions when performance degrades: which component is responsible, and which available action best restores task performance. These are distinct decision problems. Under discrete encoder switching, correctly identifying the dominant information-loss component does not necessarily identify the action that most reduces task error. Diagnosis and intervention should therefore be treated as separate problems. We support this finding with two complementary instruments. The first is a receiver-side diagnostic that partitions task-information loss into encoder and channel components and supplies population-valid task-error bounds conditional on one-sided information estimators. The second is an offline-supervised, online label-free action-value controller built from two monotone (isotonic) maps that target a reduction in measured error. The main experiments focus on two complementary axis-aligned actions that cleanly isolate the effect of discrete switching; a pilot further examines a combined action. Evaluation covers a controlled model with closed-form ground truth, three DeepJSCC vision systems, and a non-vision pilot. On STL-10, the diagnosis correctly identifies the larger loss component at 80% of operating points, yet matches the oracle-best action only 25% of the time. Structural mismatch caps any diagnosis-driven loss-ordering policy at 46% accuracy, even with perfect knowledge of the ordering. The action-value controller, which exploits the magnitude of each loss deficit, achieves 0.86, compared to 0.28 for the raw rule. Where the encoder is rarely the bottleneck, the controller naturally retains the physical-layer default, yielding robust and selective intervention.
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-11 22:07 · 最新观测版本
When Cause Attribution Is Not Enough: Action-Value Control for Task-Oriented Semantic Communication with Discrete Encoder Switching
A task-oriented semantic communication receiver that selects among independently trained encoders must answer two questions when performance degrades: which component is responsible, and which available action best restores task performance. These are distinct decision problems. Under discrete encoder switching, correctly identifying the dominant information-loss component does not necessarily identify the action that most reduces task error. Diagnosis and intervention should therefore be treated as separate problems. We support this finding with two complementary instruments. The first is a receiver-side diagnostic that partitions task-information loss into encoder and channel components and supplies population-valid task-error bounds conditional on one-sided information estimators. The second is an offline-supervised, online label-free action-value controller built from two monotone (isotonic) maps that target a reduction in measured error. The main experiments focus on two complementary axis-aligned actions that cleanly isolate the effect of discrete switching; a pilot further examines a combined action. Evaluation covers a controlled model with closed-form ground truth, three DeepJSCC vision systems, and a non-vision pilot. On STL-10, the diagnosis correctly identifies the larger loss component at 80% of operating points, yet matches the oracle-best action only 25% of the time. Structural mismatch caps any diagnosis-driven loss-ordering policy at 46% accuracy, even with perfect knowledge of the ordering. The action-value controller, which exploits the magnitude of each loss deficit, achieves 0.86, compared to 0.28 for the raw rule. Where the encoder is rarely the bottleneck, the controller naturally retains the physical-layer default, yielding robust and selective intervention.
查看此版本解读 →查看此版本原文 ↗