INDEX

Explanations

phrases related to technology and programming

oai_token-act-pair · gpt-3.5-turbo

product and brand names related to various domains and contexts

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

IFrame

Link

•Induction Feature

No Comments

Negative Logits

ãĤ¶

-0.56

ivated

-0.56

kef

-0.54

igl

-0.52

 guiActiveUn

-0.51

tun

-0.50

ppa

-0.50

bris

-0.49

lb

-0.48

elight

-0.47

POSITIVE LOGITS

 Profile

0.56

 refers

0.55

 minus

0.55

 fixme

0.53

 consists

0.51

 Location

0.50

,,,,

0.49

 itself

0.49

 implies

0.49

 Anon

0.48

Activations Density 0.736%

phrases related to technology and programming

product and brand names related to various domains and contexts

No Comments

No Known Activations