INDEX

Explanations

mentions of TV news broadcasters

oai_token-act-pair · gpt-3.5-turbo

references to television and broadcasting entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

currency

-0.71

urion

-0.70

hammad

-0.68

cule

-0.68

lins

-0.66

ventory

-0.66

hot

-0.64

ven

-0.63

leans

-0.63

POSITIVE LOGITS

ILLE

1.39

NZ

0.93

 Everywhere

0.91

 Channel

0.85

ideo

0.81

 Camer

0.76

Sav

0.75

 Broadcasting

0.74

oggles

0.73

LV

0.73

Activations Density 0.033%

mentions of TV news broadcasters

references to television and broadcasting entities

No Comments

No Known Activations