INDEX

Explanations

phrases related to a prominent figure named "Amanda" or variations like "Mama"

oai_token-act-pair · gpt-3.5-turbo

instances of the name "ama" and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ting

-0.88

tle

-0.79

points

-0.78

ted

-0.76

sheet

-0.76

ther

-0.74

ried

-0.72

lip

-0.71

teenth

-0.71

ritch

-0.68

POSITIVE LOGITS

Ã±a

1.01

Ã±

0.98

qua

0.95

ican

0.94

udeau

0.88

ama

0.87

uthor

0.85

utra

0.84

éĹĺ

0.82

icans

0.81

Activations Density 0.015%

phrases related to a prominent figure named "Amanda" or variations like "Mama"

instances of the name "ama" and related terms

No Comments

No Known Activations