arXiv cs.CL· Marcelo Samoilenko·· 11 小时前AI 评分20
仅哈希发送端能走多远？短文本消息发送端内存与比特数的预注册测绘
How far does a hash-only sender go? A pre-registered map of sender memory versus bits for short text messages
AI 导读
一项预注册实验在 Taskmaster-1 的 1,000 条英文用户话语上对比仅计算哈希的发送端与域内 PPM order-1 发送端：前者首次尝试送达 857 条，每条消息成本 192.8 bits，比 PPM order 1 多 18.4 bits（95% bootstrap 区间 10.9 至 26.3），按预注册规则判定为花费更高且不在一字节内等价。
正文 · 原文
Abstract:Can a sender that only computes hashes, with a language model only at the receiver, transmit short human text more cheaply than a classical sender trained on the same domain? We answer with one pre-registered run on 1,000 English user utterances from Taskmaster-1. The primary endpoint is the cost per message to deliver 100% of the messages (hash bits, plus a 2-byte negative acknowledgement and the raw text when the first attempt fails), paired against an in-domain PPM order-1 sender. The hash-only sender delivered 857 of 1,000 messages on the first attempt and cost 192.8 bits per message, 18.4 bits more than PPM order 1 (95% bootstrap interval 10.9 to 26.3); by the pre-registered rule it spends more and is not equivalent within one byte. As secondary results we map measured sender memory against bits. The hash-only sender (8,190 B on a host build, keyed BLAKE2b tag included, as in every method) was the cheapest point in bits from 8,190 B up to, but excluding, PPM order 1 at 24,844 B, 45 to 46 bits below smaz2 and Unishox2; from PPM order 1 upwards, in-domain senders spend fewer bits, except order 2 pruned to 16 KiB. In airtime it loses to PPM order 1 on all five links priced, to Unishox2 on four, and to raw text with a tag on two, because each failure costs an acknowledgement frame and a fallback frame. A wider receiver search recovers half of the failures and makes the cost 7.7 bits lower than PPM order 1, at about three minutes of receiver time per failure, and still loses in airtime on four of five links. We report the measurement as a delimitation, with every deviation of the run.
| Comments: | 10 pages, 1 figure, 2 tables. Code, data and per-message results: this https URL |
| Subjects: | Information Theory (cs.IT); Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.09848 [cs.IT] |
| (or arXiv:2611.09848v1 [cs.IT] for this version) | |
| https://doi.org/10.48550/arXiv.2611.09848 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Marcelo Samoilenko [view email]
[v1]
Fri, 6 Nov 2026 18:31:24 UTC (13 KB)
来源：arXiv cs.CL · arxiv.org