AI 与机器学习2024arxiv.org
Blending Is All You Need：更便宜、更优的万亿参数 LLM 替代方案
Blending Is All You Need: Cheaper, Better Alternative to Trillion-Parameters LLM
入门 · 约 30 分钟读完前置知识：大语言模型基础、聊天机器人评估方法
导读
这篇论文提出一种“混合”方法：不依赖单个大参数量 LLM，而是让多个较小的聊天 AI 轮流随机生成回复（每个模型的回复以之前其他模型生成的回复为条件），声称这样能以更低成本达到接近甚至超过超大模型的效果。评估上作者用用户留存率和参与度等交互统计作为质量代理指标，而非人工评估。
HN 关注点：用多个小模型混合能否真的替代超大模型，以及论文用用户参与度代替人工评估的做法是否可靠。
HN 评论精选
这几条评论里：支持 1 质疑 3
支持混合方法可扩展到 LLM 之外的专用系统，潜力很大
就这样？居然有效？挺有用的。这个方法似乎并不要求所有聊天 AI 都是 LLM——可以扩展到专用系统：求解器或知识库（如 Wolfram Alpha 或数据库前端）、能完成简单任务的 Alexa/Siri 级别系统。带自然语言输入输出的领域专用系统已经存在几十年了。
Animats查看英文原文
"Responses are selected randomly from a group of base chat AIs. ... The response generated by a specific chat AI is conditional on all previous responses generated by the previously selected chat AIs." That's all? That works? Useful. Could that be extended? It doesn't seem inherent in this that all the chat AIs have to be LLMs. Some might be special-purpose systems. Solvers or knowledge bases, such as Wolfram Alpha or a database front end, could play too. Systems at the Alexa/Siri level that can do simple tasks. Domain-specific systems with natural language in and out have been around for decades.
质疑质疑论文用参与度指标替代人工评估，缺乏“更好”的证据
这篇论文基本站不住脚，因为他们从未真正提供“更好”的证据。他们用用户留存率和参与度作为质量代理指标，理由是人工评估成本高。也许留存/参与度确实与人工评估相关，但你应该两者都做，先证明强相关，再决定用廉价代理指标替代人工评估。而且考虑到聊天 LLM 的已知问题，还应检查：模型显得“吸引人”是否只是因为用户要反复调整提示词？回复是否包含大量吸引人但不真实的幻觉？连续回复之间的一致性是否下降，反而意外地维持了参与度？总之，认为和这些聊天机器人对话不浪费用户时间、读者看这篇论文也不浪费时间，这种态度很草率。
abeppu查看英文原文
Ok, this seems bunk basically because they never really provide evidence of "better". > ... traditiontal gold-standard approaches use human evaluators that score the quality of generated responses, which can be costly. However, since chat AIs are by definition deployed in social environments with humans, one can leverage statistics of users interaction as a meaningful and aligned measure of chat AI engagingness and quality. To assess the ’quality’ of a chat AI, we consider two main proxy functions: the industry standard user retention and the main objective function, user engagement. Maybe retention and engagement _are_ sufficiently well correlated to human evaluations, but you should probably do both and show that they're strongly correlated before you decide to just drop the human evaluators in favor of your cheap proxy measurements. And in this field, where there are some known issues with chat LLMs, perhaps it's important to check stuff like: - Does the model seem "engaging" just b/c the user has to refine their prompt several times before they get a satisfying response? - Do responses include a lot of hallucinations which might be engaging but not true? - Do successive responses show decreased consistency or coherence between messages, in a way that might accidentally elicit continued engagement? Overall, it seems sloppy to believe that it's not a waste of humans time to talk to your chatbots, and it's not a waste of time for readers to look at this paper about your chatbots, but it's too expensive for you to actually measure the quality of responses from your chatbots.
质疑从工程角度质疑混合方法的实际计算开销
现在想想，这种“技术”是否会让每个生成 token 的计算和内存开销变成三倍？因为每个模型还要为它没生成过、从未见过的前两个 token 计算和存储 KV 值。编辑：再想想，取决于具体实现，另外两个 token 可能是并行处理的，所以应该不会慢太多。
miven查看英文原文
Now that I think about it, doesn't this "technique" triple the amount of compute and memory per generated token since each model needs to also compute and store the KV values for the two previous tokens it didn't generate and thus has never seen? Edit: On second thought, depending on how it's actually implemented the other two tokens are probably ran through the model in parallel so it shouldn't be all that much slower.
质疑调侃论文标题蹭 "All You Need" 的套路
我说过好几次了，还想再说一次——标题里带 "All You Need" 的论文，绝对不是我们真正需要的。
enoch2090查看英文原文
I've said this a few times previously yet I certainly want to say it again - "All You Need" titles are definetely not what all we need.
历次讨论
- 2024Hacker News 讨论帖147 分 · 45 评
同领域还值得读
DeepSeek-R1：通过强化学习激励 LLM 的推理能力
DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via RL
研究如何用强化学习激励大语言模型的推理能力，提出 DeepSeek-R1，并探讨 RL 训练对模型推理表现的影响。
AI 与机器学习2025 · 1,351 分 · 1,056 评 · 引用 30进阶 · 约 80 分钟编辑精选1-bit LLM 时代：面向低成本计算的三值参数
The Era of 1-bit LLMs: ternary parameters for cost-effective computing
提出用三值参数 {-1, 0, 1} 替代 LLM 中的浮点权重，即所谓“1-bit LLM”。
AI 与机器学习2024 · 1,040 分 · 447 评 · 引用 19进阶 · 约 45 分钟编辑精选DeepSeek-V3.2：推进开放大语言模型的前沿
DeepSeek-v3.2: Pushing the frontier of open large language models
这是 DeepSeek-V3.2 的技术报告，延续了 DeepSeek 开源权重模型的路线，目标是让开放模型继续逼近甚至追平前沿闭源模型。
AI 与机器学习2025 · 982 分 · 465 评进阶 · 约 90 分钟编辑精选通过自我对弈与通用强化学习掌握国际象棋和将棋（AlphaZero）
Mastering Chess and Shogi by Self-Play with General Reinforcement Learning
DeepMind 的 AlphaZero 论文：用通用强化学习算法（自我博弈 + MCTS + 神经网络），从零开始仅数小时训练便在国际象棋和将棋上击败了世界顶级引擎 Stock…
AI 与机器学习2017 · 539 分 · 270 评 · 引用 13进阶 · 约 90 分钟编辑精选用强化学习发现更快的矩阵乘法算法
Discovering faster matrix multiplication algorithms with reinforcement learning
论文提出 AlphaTensor，用强化学习搜索矩阵乘法算法。
AI 与机器学习2022 · 518 分 · 111 评专业 · 约 90 分钟编辑精选