1
CompactGUI：用 Windows 原生压缩功能，为程序和游戏文件夹瘦身
如果你的硬盘不太够用了，又不想换硬盘，那么可以试试用 CompactGUI 来救急，它可以帮你压缩现有文件。比如青小蛙用它压缩了 Epic 游戏文件夹，立省11GB空间。@Appinn 但问题也来了：压缩后会不会占 CPU？会不会影响硬盘读写？游戏还能直接运行吗？ CompactGUI 感谢 @Wι
2
Thinking at the Right Size: Amortized Distillation Across Post-Trained LLMs
不同尺寸模型间摊销蒸馏，让推理成本与思考深度灵活匹配，EMNLP 2026 前沿成果
arXiv:2608.22854v1 Announce Type: new Abstract: Practical deployment of large language models (LLMs) requires families of post-trained variants---inst…
3
Activation-Weighted Seeded Residual Coding for Low-Bit LLM Weight Repair
一种结合激活加权与残差编码的低比特LLM权重修复新法，为压缩模型性能恢复提供创新思路。
arXiv:2608.23144v1 Announce Type: new Abstract: Low-bit weight quantization saves storage but leaves errors that degrade language-model quality. We in…
4
The Compaction Cliff in Long-Running AI Agent Memory
安全规则与日志争抢上下文，压缩时规则一字之差即失效，长跑AI智能体记忆悬崖的实证研究
arXiv:2608.22752v1 Announce Type: new Abstract: A safety rule and an episodic log compete for the same tokens in an AI agent's context. When the budge…
5
SplitLite: Low-Rank Residual Compression for Split Learning
用低秩残差压缩给分割学习“减负”，通信开销大减，隐私与效率兼得的新思路。
arXiv:2608.23018v1 Announce Type: cross Abstract: Federated fine-tuning of on-device large language models (LLMs) faces a significant computing burden…
6
新版《车用压缩天然气》强制性国标发布：明确甲烷值不小于 65，与欧洲标准一致
IT之家 8 月 25 日消息，据国家市场监管总局官网 8 月 24 日公告，市场监管总局（国家标准委）批准发布《车用压缩天然气》（GB 18047—2026）强制性国家标准， 将自 2027 年 8 月 1 日起实施 。 本次标准修订旨在提升产品质量、保障使用安全、促进贸易公平。IT之家附具体内容…
7
AgentOCR: Reimagining Agent History via Optical Self-Compression
智能体历史记录膨胀是部署大痛点，AgentOCR用光学自压缩给记忆瘦身，值得一看。
arXiv:2601.04786v3 Announce Type: replace-cross Abstract: Recent advances in large language models (LLMs) enable agentic systems trained with reinforc…
8
SCOPE: A Generative Approach for LLM Prompt Compression
LLM提示词压缩新思路：生成式方法SCOPE，有效压缩上下文同时保留性能，COLM 2026收录。
arXiv:2508.15813v2 Announce Type: replace-cross Abstract: A big issue in modern LLM applications is they tend to feed long context to LLM, which resul…
9
Alphacool 带来 ES Chiller 700 液冷压缩机：700W 解热，0.3℃ 控温
IT之家 8 月 21 日消息，Alphacool 当地时间昨日宣布 面向采用分体式液冷的工作站与服务器系统 推出 ES Chiller 700 压缩机 / 冷水机。这一设备可提供 700W 的解热能力，具备 0.3℃ 的控温精度。 ES Chiller 700 采用标准 G1/4" 螺纹进出水口，…
10
宝马副总裁吴燕彦：行业应共同抵制压缩验证周期、测试中偷工减料的浮躁行为
宝马副总裁公开呼吁行业坚守安全底线，抵制测试偷工减料，给狂飙的新能源车市敲响警钟。
IT之家 8 月 20 日消息，据“睿见 Economy”报道，车百国际主办的新能源汽车全球合作发展论坛（GNEV2026）19-21 日在上海召开，宝马集团副总裁、大中华区政府和涉外事务吴燕彦在演讲中向行业发出呼吁：牢牢树立安全底线， 共同抵制压缩验证周期、在测试中偷工减料的浮躁行为，反对将不完善…
11
SchurQuant: Groupwise Discrete Optimization for Layer-Wise LLM Quantization
大模型2-3比特量化精度崩坏？这项研究用分组离散优化突破极限，无需反向传播即可部署。
arXiv:2608.15567v1 Announce Type: new Abstract: Weight-only post-training quantization (PTQ) enables the deployment of large language models under tig…
12
The Distributional View of Knowledge Distillation
用分布视角重新拆解知识蒸馏，带你理解模型压缩背后的数学本质，适合深度学习进阶读者。
arXiv:2608.15215v1 Announce Type: cross Abstract: Token-level knowledge distillation (KD) matches two conditional distributions per position, yet the …
13
创客把 2.9MB 歌曲压缩至 21KB，用 8 个二维码打印到纸上保存
Meta开源的EnCodec神经音频编解码器，将2.9MB歌曲压至21KB，配合二维码打印实现实体化保存。
IT之家 8 月 16 日消息，据 Tom's Hardware 今日报道，一位名为 Makestreme 的创作者展示了一种特殊的“纸质磁带”：他将一首时长约 2 分钟、大小约 2.9MB 的 MP3 歌曲压缩至约 21KB，再拆分成 8 个二维码打印到纸张两面。 整张纸因此可以保存完整歌曲，但播…
14
Unifying Depth and Width Pruning for LLMs via Binary Knapsack Optimization
把LLM剪枝难题化为背包问题，SNIPER统一深度+宽度剪枝，精准控预算！
arXiv:2608.12953v1 Announce Type: cross Abstract: Structured pruning is a promising approach for compressing large language models (LLMs), yet existin…
15
5 MCP Pain Points Every Developer Hits (And How a 50KB CLI Fixes Them)
MCP上下文被工具定义吃光？实测91% token压缩方案，50KB CLI解决五大痛点，AI Agent开发必看
5 MCP Pain Points Every Developer Hits (And How a 50KB CLI Fixes Them) I've been using MCP servers with Claude Code, Cursor, and Codex for months. Eve…
16
Diffuse to Compress: Leveraging Diffusion LMs for Lossless Compression
把扩散语言模型变成无损压缩器，用生成式建模重构信息论极限，一篇脑洞大开的交叉前沿研究。
arXiv:2608.11249v1 Announce Type: cross Abstract: We study the problem of lossless text compression, motivated by the rapid growth in the collection a…
17
Benchmarking Trustworthiness of SLMs: Pre-trained vs. Compressed
预训练与压缩小模型谁更可信？IJCNN权威基准测试给出对比答案。
arXiv:2608.11981v1 Announce Type: new Abstract: Small Language Models (SLMs) have emerged as a more efficient alternative to traditional Large Languag…
18
Lost in Compaction: Evaluating Side-Constraint Loss under Context Compaction
当上下文被压缩，模型还能守住约束吗？这篇论文系统评估“侧约束损失”，揭示压缩带来的隐藏风险。
arXiv:2608.11242v1 Announce Type: cross Abstract: When the context window is under pressure, LLM systems compact prior context to continue ongoing tas…
19
Lossless codec for AI agent messages – 36% fewer tokens, overhead counted
无损压缩AI智能体消息，token直降36%，成本优化利器，开源可即用。
Article URL: https://github.com/reh8n/a2acompress Comments URL: https://news.ycombinator.com/item?id=49281816 Points: 4 # Comments: 0
20
Statistically-Lossless Quantization of Large Language Models
突破量化 trade-off：统计无损新方法，兼顾模型保真与推理加速，大模型部署必读。
arXiv:2605.02404v2 Announce Type: replace Abstract: Model quantization has become essential for efficient large language model deployment, yet existin…