INDEX

Explanations

references to the name "Sig" in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to the name "Sig" or related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

McH

-0.77

ģĸ

-0.70

 Cassidy

-0.70

 GOODMAN

-0.68

ĺħ

-0.66

assetsadobe

-0.65

iard

-0.64

avez

-0.64

ĸļ

-0.63

POSITIVE LOGITS

Sig

0.98

ldom

0.90

lasses

0.89

ourney

0.89

urd

0.89

lar

0.84

ram

0.84

wered

0.83

ging

0.82

gers

0.81

Activations Density 0.019%

references to the name "Sig" in various contexts

references to the name "Sig" or related terms

No Comments

No Known Activations