1
Show HN: Latent-free ternary LLM training
无需隐变量，三值量化LLM训练新方法，开源项目BitBop让你以极低成本体验高效训练。
Article URL: https://github.com/ValerioDolci/bitbop Comments URL: https://news.ycombinator.com/item?id=48892013 Points: 1 # Comments: 1
2
Reducing HBM Bottlenecks in JAX-Based LLM Training with Host Offloading
NVIDIA官方详解如何利用主机卸载技术，在JAX中缓解HBM瓶颈，支持更大规模LLM训练。
Article URL: https://developer.nvidia.com/blog/reducing-high-bandwidth-memory-bottlenecks-in-jax-based-llm-training-with-host-offloading/ Comments URL…
3
Train Smarter, Not Longer: Memorization-Guided Data Reuse for Efficient LLM Training
教你用记忆化指导数据重用的策略，在不大幅延长训练时间的前提下提升LLM训练效率，顶级会议论文的实用思路。
arXiv:2607.04969v1 Announce Type: new Abstract: The training paradigm of large language models has shifted from traditional one-pass training to multi…
4
DeadPool: Resilient LLM Training with Hot-Swapping via Zero-Overhead Checkpoint
零开销检查点热替换技术，让大模型训练可无缝应对软硬件故障，大幅提升训练韧性
arXiv:2607.01646v1 Announce Type: new Abstract: State-of-the-art large language model (LLM) training takes tens of thousands of graphics processing un…
5
UATC – A Closed-Loop Controller to Prevent GPU OOM During LLM Training
融合工业控制理论，用卡尔曼滤波+PID+史密斯预测器精准防止LLM训练时GPU OOM，开源实现值得一试。
Article URL: https://github.com/sajjaddoda72-design/UATC Comments URL: https://news.ycombinator.com/item?id=48734992 Points: 2 # Comments: 0
6
On the Nonlinearity of Learning Rate Scaling for LLM Training
揭秘LLM训练中学习率缩放的非线性规律，为优化大模型训练过程提供全新理论视角。
arXiv:2606.29158v1 Announce Type: new Abstract: Learning-rate transfer can reduce the cost of training large language models: instead of sweeping lear…
7
Geometrically Principled Randomized Optimization for Efficient LLM Training
基于几何原理的随机优化方法，为大规模语言模型训练效率提供新思路
arXiv:2510.01878v2 Announce Type: replace Abstract: Low-rank gradient optimization for large language models is currently divided into two categories:…
8
Mechanism-Driven Monitors for Preemptive Detection of LLM Training Instability
提出机制驱动监控器，抢先检测LLM训练不稳定，提升大模型训练可靠性
arXiv:2606.28116v1 Announce Type: new Abstract: Frontier large language model training consumes massive accelerator fleets and long wall-clock computa…
9
FORGE: Fused On-Register Gradient Elimination for Memory-Efficient LLM Training
融合反向传播与优化器阶段的梯度处理，大幅降低LLM训练的内存峰值，突破显存瓶颈的新方案。
arXiv:2606.22932v1 Announce Type: new Abstract: Reverse-mode differentiation computes every weight gradient, writes it to memory, and only then lets t…
10
Online Dynamic Batching with Formal Guarantees for LLM Training
形式化保证了LLM训练中在线动态批处理的高效性与稳定性，理论扎实实验详尽。
arXiv:2606.19989v1 Announce Type: cross Abstract: Modern LLM training breaks a core assumption behind offline batch samplers: the true training cost o…
11
BLADE: Scalable Bi-level Adaptive Data Selection for LLM Training
双层次自适应数据选择方法，大幅提升大模型训练效率与数据质量
arXiv:2606.18650v1 Announce Type: new Abstract: As Large Language Model (LLM) datasets scale to trillions of tokens, data selection has emerged as a c…
12
The Curse and Blessing of Mean Bias in FP4-Quantized LLM Training
FP4训练中均值偏差的双重效应，揭示极端激活值如何破坏长尾信号，提出关键改进方向
arXiv:2603.10444v2 Announce Type: replace-cross Abstract: FP4 training promises substantial memory and compute savings for large language models, but …
13
Tweaking GPU Clock Frequency Cuts LLM Training Energy
调整GPU时钟频率，可在不影响性能前提下节省LLM训练能耗高达14%。
Article URL: https://spectrum.ieee.org/llm-training-energy-saving-trick Comments URL: https://news.ycombinator.com/item?id=48478848 Points: 3 # Commen…
14
POET-X: Memory-efficient LLM Training by Scaling Orthogonal Transformation
ICML 2026 Oral论文，提出通过扩展正交变换实现大模型训练的内存高效方案。
arXiv:2603.05500v2 Announce Type: replace Abstract: Efficient and stable training of large language models (LLMs) remains a core challenge in modern m…
15
Training an LLM in Swift, Part 2: macOS built-in frameworks
用Swift在macOS上复现GPT-2，深度揭秘vImage与BLAS底层加速技巧。
Article URL: https://www.cocoawithlove.com/blog/macos-ml-frameworks.html Comments URL: https://news.ycombinator.com/item?id=48442089 Points: 2 # Comme…
16
Better Literary Translation: A Multi-Aspect Data Generation and LLM Training Approach
文学翻译高质量数据稀缺？新框架用多维度迭代生成参考与偏好数据，提升LLM翻译流畅性与文学效果。
arXiv:2606.05924v1 Announce Type: cross Abstract: Literary translation poses unique challenges due to the scarcity of high-quality annotated data and …
17
RL Excursions during Pre-Training: Re-examining Policy Optimization for LLM training
预训练阶段引入强化学习探索，重新审视LLM策略优化方法，带来新训练范式视角。
arXiv:2606.04272v1 Announce Type: new Abstract: The standard LLM training pipeline applies reinforcement learning (RL) only after pre-training and sup…
18
Train your own LLM? Here's what happens
从零训练一个大语言模型有多难？作者分享端到端构建LLM的真实过程与关键教训。
Article URL: https://www.exasol.com/blog/train-your-own-llm/ Comments URL: https://news.ycombinator.com/item?id=48396068 Points: 7 # Comments: 0
19
EvoTrainer: Co-Evolving LLM Policies and Training Harnesses for Autonomous Agentic Reinforcement Learning
共进化策略与训练框架，突破传统静态训练瓶颈，实现LLM自主代理强化学习新范式。
arXiv:2606.03108v1 Announce Type: new Abstract: Autonomous LLM training is often framed as recipe search, which leaves the training harness largely st…
20
Memory-Efficient LLM Training with Dynamic Sparsity: From Stability to Practical Scaling
ICML2026接收：动态稀疏性突破LLM训练内存瓶颈，兼顾稳定性与实用规模化
arXiv:2606.00888v1 Announce Type: new Abstract: Dynamic Sparse Training (DST) offers a promising paradigm for improving the training and inference eff…