INDEX

Explanations

items to be added or additional information

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Add" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 stationary

-0.72

 prevailed

-0.70

Ñı

-0.68

arily

-0.66

itsu

-0.65

tarian

-0.65

 originated

-0.64

Ïĥ

-0.64

 defended

-0.63

hog

-0.63

POSITIVE LOGITS

Add

3.46

Add

2.65

add

1.85

add

1.77

ADD

1.72

 Adding

1.69

ADD

1.43

Adding

1.40

addons

1.33

 Adds

1.32

Activations Density 0.019%

items to be added or additional information

instances of the word "Add" in various contexts

No Comments

No Known Activations