INDEX

Explanations

adjectives indicating size or scale

oai_token-act-pair · gpt-3.5-turbo

words that denote significant or extreme quantities and conditions

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

KO

-0.76

orthy

-0.75

arate

-0.74

ONSORED

-0.72

otonin

-0.71

utics

-0.70

ertodd

-0.69

udder

-0.69

VERTISEMENT

-0.68

à¼

-0.67

POSITIVE LOGITS

est

1.17

ness

1.08

liest

0.98

 nature

0.95

 tendency

0.82

iest

0.82

nesses

0.81

 portion

0.78

 aspect

0.78

hest

0.78

Activations Density 0.416%

adjectives indicating size or scale

words that denote significant or extreme quantities and conditions

No Comments

No Known Activations