Metraipolis › archive › dc_daa9596dbace

Reading Less While Writing: A Closed-Form Bandwidth Dial for Streaming Multimodal Decoders — arXiv:2609.20845

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.20845
License (as archived): CC0-1.0
Archived: 2026-10-10T01:27:03.556Z
SHA-256 of the archived text: efa9dbf1178daed59c68f5c4a660f2770bc6f12f7430f6cd1de75c7f2077ed61
Stable URL of this page: https://metraipolis.com/documento/dc_daa9596dbace

How to cite

Reading Less While Writing: A Closed-Form Bandwidth Dial for Streaming Multimodal Decoders — arXiv:2609.20845 — arXiv (en), https://arxiv.org/abs/2609.20845 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_daa9596dbace (Metraipolis archive dc_daa9596dbace, sha256 efa9dbf1178daed59c68f5c4a660f2770bc6f12f7430f6cd1de75c7f2077ed61)

BibTeX

@misc{dcdaa9596dbace,
  title = {Reading Less While Writing: A Closed-Form Bandwidth Dial for Streaming Multimodal Decoders — arXiv:2609.20845},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.20845},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_daa9596dbace; sha256 efa9dbf1178daed59c68f5c4a660f2770bc6f12f7430f6cd1de75c7f2077ed61},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.20845. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

A decoder that turns video or audio into text conventionally consumes the entire input before emitting a word. Offline this is merely more than the task requires; live it is impossible, since a caption cannot wait for a match to end. Streaming systems bolt on a fixed rule such as wait-$k$, which waits for the same number of input tokens before every word, regardless of the input's length or pace. We replace the fixed offset with ZENDAYA, a schedule governed by a single continuous parameter $\gamma$. It makes the visible source prefix a closed-form function of generation progress, scaled by the input's own predicted length, so an ordinary offline decoder and a real-time streaming decoder become two endpoints of one family rather than separate models. The same scalar fixes, in closed form, the mean fraction of source consumed per emitted word, $\bar{E}(\gamma) \approx 1/(1+\gamma)$, making it at once a latency dial and an interpretable budget. We prove a structural dependency theorem: under any schedule fixed in advance and non-decreasing, no emitted token can depend on input that has not yet arrived. The guarantee holds for trained and untrained weights alike, and extends to unbounded streams under arbitrary asynchronous arrival. The empirical result is counterintuitive: seeing less can produce better text, because a flood of source dilutes attention exactly when the model has the least of its own output to anchor on. Trained from scratch across two modalities and three public corpora (Charades-STA, ActivityNet Captions, LibriHeavy), a compact 29M-parameter decoder matches or beats the fixed schedule while reading less of the source, with the sharpest gains at the lowest latencies, where a fixed offset collapses. Streaming METEOR gains are statistically significant on all three corpora.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-21 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.20845 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-10T01:00:30.255Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).