INDEX

Explanations

words related to a direction or place, specifically referring to the "back."

oai_token-act-pair · gpt-3.5-turbo

instances of the word "back" 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 @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 fuss

-0.65

 Aires

-0.62

tein

-0.61

mble

-0.61

BUS

-0.60

 faint

-0.59

Osc

-0.59

urious

-0.56

IFIED

-0.56

 LINE

-0.54

POSITIVE LOGITS

door

1.17

back

1.11

wards

1.06

lash

1.06

dated

1.01

gam

0.94

tracking

0.92

)=(

0.91

doors

0.91

hoe

0.90

Activations Density 0.015%

words related to a direction or place, specifically referring to the "back."

instances of the word "back" in various contexts

No Comments

No Known Activations