π |
AI |
Nature |
AI μμ±λ¬Όλ‘ λ€μμ±μ 0μΌλ‘ μλ ΄νλ€. (2024λ
7μ Nature λ
Όλ¬Έ)
[Youtube]
|
ACL 2024 |
GPT λͺ¨λΈμ λλ€ν μ§μμ μ κ·Όν μ μλ€.
[Youtube]
|
ICLR 2023 |
격μμΈν¬, μ₯μμΈν¬, and Blind Agent
[Youtube]
|
Nature |
μΈκ³΅μ§λ₯μ λ§μμ΄λ‘ (Theory of Mind in 2024 )
[Youtube]
|
π |
Machine Learning |
|
Jensen's Inequality
[Youtube]
|
|
Evidence Lower Bound (ELBO), VAE
[Youtube]
|
π |
LLMs |
|
Prover-Verifier Games Improve Legibility of LLM Outputs
[TBD]
|
|
Probabilistic Inference for LLM (ICML 2024 Best Paperλ₯Ό μ½κΈ° μν μ€λΉ)
[Youtube]
|
π |
Mechanistic Interpretability |
Anthropic |
Transformer Induction Head
[Youtube]
|
Anthropic |
Scaling Monosemanticity 4/5, UMAP
[Youtube]
|
ICML 2024 |
In-context Learning and In-context Vector
[Youtube]
|
ICML 2024 WS |
Implicit Vocabulary in GPT
[Youtube]
|
ICML 2024 WS |
Dictionary Learning for Language Model Interpretability with Board Game Models
[OpenReview]
|
ICML 2024 WS |
The Geometry of Categorical and Hierarchical Concepts
[OpenReview]
|
ICML 2024 |
Model Surgery: Modulating LLM's Behavior Via Simple Parameter Editing
[Youtube]
|
π |
Causality |
NeurIPS 2022 |
Indirect Effect and Locating Factuality in GPT
[Youtube]
|
CLR |
Distributed Alignment Search (Intuition)
[Youtube]
|
CLR |
Distributed Alignment Search (Detailed Mathematical Formulation)
[Youtube]
|
CLR |
Distributed Alignment Search (Philosophy)
[Youtube]
|
CLR |
Distributed Alignment Search (Easy Math)
[Youtube TBD]
|
NeurIPS 2023 |
Distributed Alignment Search (Application with Alpaca)
[Youtube TBD]
|
π |
Quantization |
|
The Era of 1-bit LLMs: All Large Language Models are in 1.58 Bits
[Youtube]
|
π |
Diffusion |
|
Anomaly Detection with Diffusion
[Youtube]
|
|
Rectified Flow (Flow Straight and Fast)
[Youtube]
|