INDEX

Explanations

names or terms related to the name "Sig"

oai_token-act-pair · gpt-3.5-turbo

the presence of the word "Sig" 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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĺħ

-0.73

McH

-0.72

 Cassidy

-0.68

ĸļ

-0.63

assetsadobe

-0.61

ģĸ

-0.61

avez

-0.60

TTC

-0.59

uclear

-0.59

iard

-0.59

POSITIVE LOGITS

lasses

0.92

Sig

0.91

urd

0.89

ging

0.87

ldom

0.86

gers

0.85

ourney

0.85

wered

0.84

ifts

0.82

ificant

0.82

Activations Density 0.014%

names or terms related to the name "Sig"

the presence of the word "Sig" in various contexts

No Comments

No Known Activations