INDEX

Explanations

locations or places

oai_token-act-pair · gpt-3.5-turbo

phrases indicating choices or alternatives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ires

-0.69

istar

-0.65

moil

-0.64

Ident

-0.64

mitter

-0.64

emo

-0.63

ETS

-0.62

hani

-0.62

EMP

-0.60

HER

-0.59

POSITIVE LOGITS

nam

1.07

chard

1.07

ifice

1.05

acle

1.00

Else

0.99

chid

0.98

nery

0.95

acles

0.93

ific

0.91

lando

0.90

Activations Density 0.150%

locations or places

phrases indicating choices or alternatives

No Comments

No Known Activations