Tag: LLM

20 items with this tag.

Jul 20, 2026
无反向传播的权重空间后训练：ZO、ES 与参数采样
- LLM
- agentic-RL
Jun 16, 2026
Sparsity Curse：从模型合并理解 RLVR 参数空间
- LLM
- agentic-RL
Jun 15, 2026
CircuitLasso：用稀疏回归做可扩展的 LLM 电路发现
- LLM
Jun 01, 2026
From Brewing to Resolution: Tracing the Internal Lifecycle of Code Reasoning in LLMs
- LLM
May 02, 2026
Reasoning emerges from constrained inference manifolds in large language models
- LLM
Jan 30, 2026
The Blessing of Dimensionality in LLM Fine-tuning: A Variance-Curvature Perspective
- LLM
- agentic-RL
Mar 27, 2025
Circuit Tracing / On the Biology of a Large Language Model
- interpretability
- LLM
Jan 27, 2025
参数空间可解释性：APD 与 SPD
- interpretability
- LLM
Jan 11, 2024
Patchscopes: A Unifying Framework for Inspecting Hidden Representations of LMs
- interpretability
- LLM
Oct 04, 2023
Towards Monosemanticity：用字典学习把语言模型解耦成单义特征
- interpretability
- LLM
Oct 02, 2023
Representation Engineering: A Top-Down Approach to AI Transparency
- interpretability
- LLM
Apr 28, 2023
ACDC：迈向机制可解释性的自动化电路发现
- interpretability
- LLM
Mar 05, 2023
DAS：在可解释因果变量与分布式神经表征之间寻找对齐
- interpretability
- LLM
Nov 01, 2022
Interpretability in the Wild：GPT-2 small 的 IOI 电路
- interpretability
- LLM
Sep 14, 2022
Toy Models of Superposition
- interpretability
- LLM
Mar 08, 2022
In-context Learning and Induction Heads
- interpretability
- LLM
Feb 10, 2022
Locating and Editing Factual Associations in GPT (ROME)
- interpretability
- LLM
Dec 22, 2020
Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tuning
- LLM
Aug 31, 2020
interpreting GPT: the logit lens
- interpretability
- LLM
Apr 24, 2018
Measuring the Intrinsic Dimension of Objective Landscapes
- LLM

if_Guo

Explorer

Tag: LLM

无反向传播的权重空间后训练：ZO、ES 与参数采样

Sparsity Curse：从模型合并理解 RLVR 参数空间

CircuitLasso：用稀疏回归做可扩展的 LLM 电路发现

From Brewing to Resolution: Tracing the Internal Lifecycle of Code Reasoning in LLMs

Reasoning emerges from constrained inference manifolds in large language models

The Blessing of Dimensionality in LLM Fine-tuning: A Variance-Curvature Perspective

Circuit Tracing / On the Biology of a Large Language Model

参数空间可解释性：APD 与 SPD

Patchscopes: A Unifying Framework for Inspecting Hidden Representations of LMs

Towards Monosemanticity：用字典学习把语言模型解耦成单义特征

Representation Engineering: A Top-Down Approach to AI Transparency

ACDC：迈向机制可解释性的自动化电路发现

DAS：在可解释因果变量与分布式神经表征之间寻找对齐

Interpretability in the Wild：GPT-2 small 的 IOI 电路

Toy Models of Superposition

In-context Learning and Induction Heads

Locating and Editing Factual Associations in GPT (ROME)

Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tuning

interpreting GPT: the logit lens

Measuring the Intrinsic Dimension of Objective Landscapes