INDEX

Explanations

the word "Fun" with various different contexts

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "Fun."

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

xual

-1.02

 Centauri

-0.72

Minimum

-0.67

ib

-0.67

 Uriel

-0.63

 Sapphire

-0.62

 Twain

-0.61

×Ļ

-0.60

 ozone

-0.59

ATTLE

-0.59

POSITIVE LOGITS

nel

1.33

nels

1.30

eral

1.29

imation

1.22

nell

1.09

ctions

1.09

stant

0.94

ctor

0.92

combe

0.91

kel

0.90

Activations Density 0.013%

the word "Fun" with various different contexts

mentions of the word "Fun."

No Comments

No Known Activations