INDEX

Explanations

questions being asked in text

oai_token-act-pair · gpt-3.5-turbo

questions and statements that begin with "Is," indicating inquiries or examinations of a topic

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ãģĮ

-0.77

è¦ļéĨĴ

-0.72

ð

-0.64

 excess

-0.61

ESE

-0.61

 lands

-0.61

 hoop

-0.60

 rites

-0.60

 depot

-0.60

IOR

-0.60

POSITIVE LOGITS

olated

1.35

olation

1.25

olate

1.23

abella

1.15

htar

1.08

ync

1.01

sei

1.00

rael

1.00

peria

0.98

terness

0.95

Activations Density 0.085%

questions being asked in text

questions and statements that begin with "Is," indicating inquiries or examinations of a topic

No Comments

No Known Activations