INDEX

Explanations

words related to performances or appearances

oai_token-act-pair · gpt-3.5-turbo

words related to awareness or attention in contexts discussing societal issues

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

amar

-0.76

rador

-0.75

 forget

-0.75

onut

-0.75

ingen

-0.73

acca

-0.71

ocular

-0.71

odor

-0.69

GER

-0.68

REM

-0.68

POSITIVE LOGITS

llor

0.94

ments

0.94

ment

0.82

ances

0.81

Ĥª

0.79

utical

0.75

utics

0.73

bringer

0.72

etus

0.72

ance

0.71

Activations Density 0.020%

words related to performances or appearances

words related to awareness or attention in contexts discussing societal issues

No Comments

No Known Activations