INDEX

Explanations

phrases related to medical conditions or surnames containing "ill"

oai_token-act-pair · gpt-3.5-turbo

references to the term "guilt."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

EStream

-0.73

 compr

-0.68

ccording

-0.64

BOOK

-0.63

lished

-0.63

resy

-0.62

 nationally

-0.62

 Debor

-0.62

 gobl

-0.62

 Stacy

-0.60

POSITIVE LOGITS

espie

1.48

uminati

1.30

iard

1.16

icit

1.09

inois

1.05

omon

1.00

ustration

1.00

inger

0.96

iam

0.96

umin

0.94

Activations Density 0.029%

phrases related to medical conditions or surnames containing "ill"

references to the term "guilt."

No Comments

No Known Activations