INDEX

Explanations

the keyword "kn," which might be an abbreviation or shorthand for various longer words

oai_token-act-pair · gpt-3.5-turbo

occurrences of the substring "kn" 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 @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

DEM

-0.84

geist

-0.79

IUM

-0.74

 Ames

-0.70

grounds

-0.70

quickShipAvailable

-0.70

Ø§ÙĦ

-0.68

Gum

-0.68

 Ducks

-0.67

 Lutheran

-0.66

POSITIVE LOGITS

uckle

1.21

itted

1.18

uckles

1.06

ickers

1.05

ighth

1.02

obb

0.97

ivable

0.97

obs

0.95

itting

0.94

idd

0.94

Activations Density 0.014%

the keyword "kn," which might be an abbreviation or shorthand for various longer words

occurrences of the substring "kn" in various contexts

No Comments

No Known Activations