会议 2022
Post-training
LoRA
LoRA: Low-Rank Adaptation of Large Language Models
ICLR 2022 · 参数高效适配的重要历史工作（DBLP核验2026-10-29） ↗ONE PAPER A DAY
每日精读Reading the classics
每天一篇旧作：从动机与背景，到关键公式与可以继续追问的问题。
已读论文
16 篇会议 2021
会议 2019
Deep learning theory
Lottery Ticket Hypothesis
The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks
ICLR 2019 Best Paper Award ↗会议 2021
会议 2015
会议 2024
RNN / linear attention
Mamba
Mamba: Linear-Time Sequence Modeling with Selective State Spaces
COLM 2024 Outstanding Paper Award ↗会议 2018
Deep learning theory
GD max-margin bias
The Implicit Bias of Gradient Descent on Separable Data
JMLR 2018 · 隐式偏置的重要历史工作 ↗会议 2020
Scaling laws
Kaplan scaling laws
Scaling Laws for Neural Language Models
Scaling laws 重要历史工作；OpenAlex 条目 1,745 次引用（查询 2026-10-22） ↗会议 2022
Post-training
InstructGPT
Training Language Models to Follow Instructions with Human Feedback
NeurIPS 2022；ACM Digital Library 于 2026-10-21 显示 33,248 次引用 ↗会议 2021
Diffusion models
Score-SDE
Score-Based Generative Modeling through Stochastic Differential Equations
ICLR 2021 Outstanding Paper Award ↗会议 2020
RNN / linear attention
Linear Transformers
Transformers are RNNs: Fast Autoregressive Transformers with Linear Attention
ICML 2020 · 线性注意力的重要历史工作 ↗会议 2018
Deep learning theory
Neural Tangent Kernel
Neural Tangent Kernel: Convergence and Generalization in Neural Networks
NeurIPS 2018 · 深度学习理论重要历史工作 ↗会议 2022
Scaling laws
Chinchilla
Training Compute-Optimal Large Language Models
NeurIPS 2022 · scaling laws 重要历史工作 ↗会议 2023
Post-training
DPO
Direct Preference Optimization: Your Language Model is Secretly a Reward Model
NeurIPS 2023 Oral · Outstanding Main Track Runner-Up ↗会议 2020
Diffusion models
DDPM
Denoising Diffusion Probabilistic Models
高引用 · OpenAlex 所查条目 5,441 次（查询 2026-10-15） ↗会议 2022
RNN / linear attention
S4
Efficiently Modeling Long Sequences with Structured State Spaces
ICLR 2022 Oral · Outstanding Paper Honorable Mention ↗优先选择高引用、会议 best paper 或 oral 工作，并给出可核验的入选依据。历史论文与当天新研究分别标注。
← 日报归档