INDEX

Explanations

phrases related to government, health, and politics

oai_token-act-pair · gpt-3.5-turbo

references to health-related issues and organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

colored

-0.79

 behaviors

-0.71

rollment

-0.71

 coloring

-0.70

Newsletter

-0.70

imize

-0.68

 honor

-0.68

Defense

-0.68

 Hawai

-0.67

 colored

-0.67

POSITIVE LOGITS

 Â£

1.20

 realise

1.17

 recognise

1.16

 centre

1.13

 apologise

1.10

 scept

1.09

 emphas

1.08

 apologised

1.08

 recognised

1.08

 aluminium

1.08

Activations Density 3.970%

phrases related to government, health, and politics

references to health-related issues and organizations

No Comments

No Known Activations