INDEX

Explanations

phrases related to the term "Bad" with varying activations

oai_token-act-pair · gpt-3.5-turbo

references to the "Badgers," specifically in relation to sports or cultural contexts

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

ĸļ

-0.91

conservancy

-0.87

theless

-0.72

izabeth

-0.71

ittee

-0.71

ournal

-0.70

ktop

-0.70

utical

-0.69

igating

-0.69

ittees

-0.69

POSITIVE LOGITS

ges

0.98

Bad

0.96

ger

0.94

ged

0.92

bad

0.91

gers

0.90

 Luck

0.88

Bad

0.86

die

0.86

ass

0.85

Activations Density 0.011%

phrases related to the term "Bad" with varying activations

references to the "Badgers," specifically in relation to sports or cultural contexts

No Comments

No Known Activations