INDEX

Explanations

phrases related to a process or action being done, particularly with an emphasis on the action happening, and the outcome or impact of the action

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

cause

-0.73

uesday

-0.60

uph

-0.59

afort

-0.59

Ã©t

-0.59

thumbnails

-0.58

Ingredients

-0.58

agra

-0.57

tu

-0.57

iazep

-0.57

POSITIVE LOGITS

we

0.89

0.87

 they

0.72

 alone

0.68

he

0.65

 however

0.65

 comes

0.62

,.

0.61

,...

0.60

 came

0.58

Activations Density 0.146%

phrases related to a process or action being done, particularly with an emphasis on the action happening, and the outcome or impact of the action

No Comments

No Known Activations