INDEX

Explanations

words related to drinking, especially alcohol consumption

oai_token-act-pair · gpt-3.5-turbo

references to "dr" or "drugs" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

eers

-0.85

SHIP

-0.76

yip

-0.71

eer

-0.64

hire

-0.64

quo

-0.63

eering

-0.63

ISM

-0.59

 Pose

-0.58

 button

-0.58

POSITIVE LOGITS

agons

1.45

inking

1.32

ifter

1.29

ifts

1.24

unk

1.22

inks

1.22

ums

1.20

unks

1.19

ift

1.18

ifting

1.17

Activations Density 0.040%

words related to drinking, especially alcohol consumption

references to "dr" or "drugs" in various contexts

No Comments

No Known Activations