INDEX

Explanations

words related to disruptions or actions that can be viewed as upsetting or unsettling

oai_token-act-pair · gpt-3.5-turbo

words and terms related to disability

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 ages

-0.64

 hell

-0.62

 formula

-0.59

Sut

-0.59

 Deer

-0.59

Osw

-0.58

 fetch

-0.58

 Kinnikuman

-0.58

glers

-0.58

 Hunt

-0.58

POSITIVE LOGITS

cipl

1.32

ruption

1.25

claimer

1.23

enfranch

1.21

cerning

1.20

comfort

1.16

qus

1.15

rupt

1.14

astrous

1.14

advant

1.13

Activations Density 0.013%

words related to disruptions or actions that can be viewed as upsetting or unsettling

words and terms related to disability

No Comments

No Known Activations