{"ID":2828304,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2512.14022","arxiv_id":"2512.14022","title":"Symbol Distributions in Semantic Communications: A Source-Channel Equilibrium Perspective","abstract":"Semantic communication systems often use an end-to-end neural network to map input data into continuous symbols. These symbols, which are essentially neural network features, usually have fixed dimensions and heavy-tailed distributions. However, due to the end-to-end training nature of the neural network encoder, the underlying reason for the symbol distribution remains underexplored. We propose a new explanation for the semantic symbol distribution: an inherent trade-off between source coding and communications. Specifically, the encoder balances two objectives: allocating power for minimum \\emph{effective codelength} (for source coding) and maximizing mutual information (for communications). We formalize this trade-off via an information-theoretic optimization framework, which yields a Student's $t$-distribution as the resulting symbol distribution. Through extensive studies on image-based semantic systems, we find that our formulation models the learned symbols and predicts how the symbol distribution's shape parameter changes with respect to (i) the use of variable-length coding and (ii) the dataset's entropy variability. Furthermore, we demonstrate how introducing a regularizer that enforces a target symbol distribution, which guides the encoder towards a target prior (e.g., Gaussian), improves training convergence and supports our hypothesis.","short_abstract":"Semantic communication systems often use an end-to-end neural network to map input data into continuous symbols. These symbols, which are essentially neural network features, usually have fixed dimensions and heavy-tailed distributions. However, due to the end-to-end training nature of the neural network encoder, the u...","url_abs":"https://arxiv.org/abs/2512.14022","url_pdf":"https://arxiv.org/pdf/2512.14022v1","authors":"[\"Hanju Yoo\",\"Dongha Choi\",\"Songkuk Kim\",\"Chan-Byoung Chae\",\"Robert W. Heath\"]","published":"2025-12-16T02:39:15Z","proceeding":"cs.IT","tasks":"[\"cs.IT\",\"eess.SP\"]","methods":"[]","has_code":false}