INDEX

Explanations

the word "girl" with a high activation value

oai_token-act-pair · gpt-3.5-turbo

the presence of the substring "ir" in words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 untreated

-0.64

Ĥª

-0.61

 Dragonbound

-0.60

er

-0.58

alter

-0.57

ļé

-0.56

ĸļ

-0.56

 comma

-0.56

 CLSID

-0.56

raised

-0.55

POSITIVE LOGITS

vana

1.41

rha

1.09

andom

1.04

ROR

0.99

mingham

0.93

ror

0.90

acial

0.90

oux

0.87

rh

0.86

gil

0.85

Activations Density 0.035%

the word "girl" with a high activation value

the presence of the substring "ir" in words

No Comments

No Known Activations