INDEX

Explanations

names of places or people with "Ad" in them

oai_token-act-pair · gpt-3.5-turbo

names of academic or professional titles and positions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

taboola

-0.73

 gears

-0.68

PF

-0.68

ĨĴ

-0.66

 magnification

-0.59

 inconven

-0.59

 partName

-0.59

natureconservancy

-0.58

 sidewalks

-0.58

cknowled

-0.58

POSITIVE LOGITS

ause

0.75

enza

0.72

lie

0.68

orno

0.68

aline

0.66

vantage

0.66

bia

0.65

ther

0.64

oli

0.63

itness

0.63

Activations Density 0.087%

names of places or people with "Ad" in them

names of academic or professional titles and positions

No Comments

No Known Activations