INDEX

Explanations

words related to political events and leaders

oai_token-act-pair · gpt-3.5-turbo

references to media events and figures

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 leveling

-0.85

cooked

-0.74

 chapter

-0.72

 chapters

-0.70

Chapter

-0.68

emonic

-0.66

\",

-0.66

cffff

-0.65

 Clicker

-0.64

)",

-0.64

POSITIVE LOGITS

 Getty

2.01

Getty

1.71

 REUTERS

1.67

AFP

1.58

 Reuters

1.55

Rex

1.37

Reuters

1.30

EPA

1.30

AFP

1.30

 Shutterstock

1.29

Activations Density 0.079%

words related to political events and leaders

references to media events and figures

No Comments

No Known Activations